publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-04-19T11:04:27Z | http://arxiv.org/pdf/2404.12788v1 | http://arxiv.org/abs/2404.12788v1 | Nacime Bouziani, Shubhi Tyagi, Joseph Fisher, Jens Lehmann, Andrea Pierleoni | 2024-04-19T11:04:27Z | cs.CL | Extracting structured information from unstructured text is critical for many
downstream NLP applications and is traditionally achieved by closed information
extraction (cIE). However, existing approaches for cIE suffer from two
limitations: (i) they are often pipelines which makes them prone to error
propagation, and/or (ii) they are restricted to sentence level which prevents
them from capturing long-range dependencies and results in expensive inference
time. We address these limitations by proposing REXEL, a highly efficient and
accurate model for the joint task of document level cIE (DocIE). REXEL performs
mention detection, entity typing, entity disambiguation, coreference resolution
and document-level relation classification in a single forward pass to yield
facts fully linked to a reference knowledge graph. It is on average 11 times
faster than competitive existing approaches in a similar setting and performs
competitively both when optimised for any of the individual subtasks and a
variety of combinations of different joint tasks, surpassing the baselines by
an average of more than 6 F1 points. The combination of speed and accuracy
makes REXEL an accurate cost-efficient system for extracting structured
information at web-scale. We also release an extension of the DocRED dataset to
enable benchmarking of future work on DocIE, which is available at
https://github.com/amazon-science/e2e-docie. | cs.CL | REXEL: An End-to-end Model for Document-Level Relation Extraction and Entity Linking |
2024-04-19T09:59:44Z | http://arxiv.org/pdf/2404.12753v1 | http://arxiv.org/abs/2404.12753v1 | Wenhao Huang, Chenghao Peng, Zhixu Li, Jiaqing Liang, Yanghua Xiao, Liqian Wen, Zulong Chen | 2024-04-19T09:59:44Z | cs.CL, cs.AI | Web automation is a significant technique that accomplishes complicated web
tasks by automating common web actions, enhancing operational efficiency, and
reducing the need for manual intervention. Traditional methods, such as
wrappers, suffer from limited adaptability and scalability when faced with a
new website. On the other hand, generative agents empowered by large language
models (LLMs) exhibit poor performance and reusability in open-world scenarios.
In this work, we introduce a crawler generation task for vertical information
web pages and the paradigm of combining LLMs with crawlers, which helps
crawlers handle diverse and changing web environments more efficiently. We
propose AutoCrawler, a two-stage framework that leverages the hierarchical
structure of HTML for progressive understanding. Through top-down and step-back
operations, AutoCrawler can learn from erroneous actions and continuously prune
HTML for better action generation. We conduct comprehensive experiments with
multiple LLMs and demonstrate the effectiveness of our framework. Resources of
this paper can be found at \url{https://github.com/EZ-hwh/AutoCrawler} | cs.CL | AutoCrawler: A Progressive Understanding Web Agent for Web Crawler Generation |
2024-04-19T09:44:51Z | http://arxiv.org/pdf/2404.12744v2 | http://arxiv.org/abs/2404.12744v2 | Pablo Biedma, Xiaoyuan Yi, Linus Huang, Maosong Sun, Xing Xie | 2024-05-10T06:09:02Z | cs.CL, cs.AI | Recent advancements in Large Language Models (LLMs) have revolutionized the
AI field but also pose potential safety and ethical risks. Deciphering LLMs'
embedded values becomes crucial for assessing and mitigating their risks.
Despite extensive investigation into LLMs' values, previous studies heavily
rely on human-oriented value systems in social sciences. Then, a natural
question arises: Do LLMs possess unique values beyond those of humans? Delving
into it, this work proposes a novel framework, ValueLex, to reconstruct LLMs'
unique value system from scratch, leveraging psychological methodologies from
human personality/value research. Based on Lexical Hypothesis, ValueLex
introduces a generative approach to elicit diverse values from 30+ LLMs,
synthesizing a taxonomy that culminates in a comprehensive value framework via
factor analysis and semantic clustering. We identify three core value
dimensions, Competence, Character, and Integrity, each with specific
subdimensions, revealing that LLMs possess a structured, albeit non-human,
value system. Based on this system, we further develop tailored projective
tests to evaluate and analyze the value inclinations of LLMs across different
model sizes, training methods, and data sources. Our framework fosters an
interdisciplinary paradigm of understanding LLMs, paving the way for future AI
alignment and regulation. | cs.CL | Beyond Human Norms: Unveiling Unique Values of Large Language Models through Interdisciplinary Approaches |
2024-04-19T09:15:07Z | http://arxiv.org/pdf/2404.12728v2 | http://arxiv.org/abs/2404.12728v2 | Chengwei Qin, Wenhan Xia, Tan Wang, Fangkai Jiao, Yuchen Hu, Bosheng Ding, Ruirui Chen, Shafiq Joty | 2024-06-23T05:18:50Z | cs.CL | Analogical reasoning is a unique ability of humans to address unfamiliar
challenges by transferring strategies from relevant past experiences. One key
finding in psychology is that compared with irrelevant past experiences,
recalling relevant ones can help humans better handle new tasks.
Coincidentally, the NLP community has also recently found that self-generating
relevant examples in the context can help large language models (LLMs) better
solve a given problem than hand-crafted prompts. However, it is yet not clear
whether relevance is the key factor eliciting such capability, i.e., can LLMs
benefit more from self-generated relevant examples than irrelevant ones? In
this work, we systematically explore whether LLMs can truly perform analogical
reasoning on a diverse set of reasoning tasks. With extensive experiments and
analysis, we show that self-generated random examples can surprisingly achieve
comparable or even better performance, e.g., 4% performance boost on GSM8K with
random biological examples. We find that the accuracy of self-generated
examples is the key factor and subsequently design two improved methods with
significantly reduced inference costs. Overall, we aim to advance a deeper
understanding of LLM analogical reasoning and hope this work stimulates further
research in the design of self-generated contexts. | cs.CL | Relevant or Random: Can LLMs Truly Perform Analogical Reasoning? |
2024-04-19T09:10:29Z | http://arxiv.org/pdf/2404.12726v2 | http://arxiv.org/abs/2404.12726v2 | Xinfeng Yuan, Siyu Yuan, Yuhan Cui, Tianhe Lin, Xintao Wang, Rui Xu, Jiangjie Chen, Deqing Yang | 2024-07-02T14:01:27Z | cs.CL | Large language models (LLMs) have demonstrated impressive performance and
spurred numerous AI applications, in which role-playing agents (RPAs) are
particularly popular, especially for fictional characters. The prerequisite for
these RPAs lies in the capability of LLMs to understand characters from
fictional works. Previous efforts have evaluated this capability via basic
classification tasks or characteristic imitation, failing to capture the
nuanced character understanding with LLMs. In this paper, we propose evaluating
LLMs' character understanding capability via the character profiling task,
i.e., summarizing character profiles from corresponding materials, a widely
adopted yet understudied practice for RPA development. Specifically, we
construct the CroSS dataset from literature experts and assess the generated
profiles by comparing ground truth references and their applicability in
downstream tasks. Our experiments, which cover various summarization methods
and LLMs, have yielded promising results. These results strongly validate the
character understanding capability of LLMs. Resources are available at
https://github.com/Joanna0123/character_profiling. | cs.CL | Evaluating Character Understanding of Large Language Models via Character Profiling from Fictional Works |
2024-04-19T09:00:05Z | http://arxiv.org/pdf/2404.12720v1 | http://arxiv.org/abs/2404.12720v1 | Yihao Ding, Kaixuan Ren, Jiabin Huang, Siwen Luo, Soyeon Caren Han | 2024-04-19T09:00:05Z | cs.CV, cs.CL | Document Question Answering (QA) presents a challenge in understanding
visually-rich documents (VRD), particularly those dominated by lengthy textual
content like research journal articles. Existing studies primarily focus on
real-world documents with sparse text, while challenges persist in
comprehending the hierarchical semantic relations among multiple pages to
locate multimodal components. To address this gap, we propose PDF-MVQA, which
is tailored for research journal articles, encompassing multiple pages and
multimodal information retrieval. Unlike traditional machine reading
comprehension (MRC) tasks, our approach aims to retrieve entire paragraphs
containing answers or visually rich document entities like tables and figures.
Our contributions include the introduction of a comprehensive PDF Document VQA
dataset, allowing the examination of semantically hierarchical layout
structures in text-dominant documents. We also present new VRD-QA frameworks
designed to grasp textual contents and relations among document layouts
simultaneously, extending page-level understanding to the entire multi-page
document. Through this work, we aim to enhance the capabilities of existing
vision-and-language models in handling challenges posed by text-dominant
documents in VRD-QA. | cs.CV | PDF-MVQA: A Dataset for Multimodal Information Retrieval in PDF-based Visual Question Answering |
2024-04-19T08:52:22Z | http://arxiv.org/pdf/2404.12715v2 | http://arxiv.org/abs/2404.12715v2 | Yichong Huang, Xiaocheng Feng, Baohang Li, Yang Xiang, Hui Wang, Bing Qin, Ting Liu | 2024-05-30T16:59:56Z | cs.CL | Large language models (LLMs) exhibit complementary strengths in various
tasks, motivating the research of LLM ensembling. However, existing work
focuses on training an extra reward model or fusion model to select or combine
all candidate answers, posing a great challenge to the generalization on unseen
data distributions. Besides, prior methods use textual responses as
communication media, ignoring the valuable information in the internal
representations. In this work, we propose a training-free ensemble framework
DeePEn, fusing the informative probability distributions yielded by different
LLMs at each decoding step. Unfortunately, the vocabulary discrepancy between
heterogeneous LLMs directly makes averaging the distributions unfeasible due to
the token misalignment. To address this challenge, DeePEn maps the probability
distribution of each model from its own probability space to a universal
relative space based on the relative representation theory, and performs
aggregation. Next, we devise a search-based inverse transformation to transform
the aggregated result back to the probability space of one of the ensembling
LLMs (main model), in order to determine the next token. We conduct extensive
experiments on ensembles of different number of LLMs, ensembles of LLMs with
different architectures, and ensembles between the LLM and the specialist
model. Experimental results show that (i) DeePEn achieves consistent
improvements across six benchmarks covering subject examination, reasoning, and
knowledge, (ii) a well-performing specialist model can benefit from a less
effective LLM through distribution fusion, and (iii) DeePEn has complementary
strengths with other ensemble methods such as voting. | cs.CL | Ensemble Learning for Heterogeneous Large Language Models with Deep Parallel Collaboration |
2024-04-19T08:45:42Z | http://arxiv.org/pdf/2404.13099v1 | http://arxiv.org/abs/2404.13099v1 | Avinash Anand, Mohit Gupta, Kritarth Prasad, Navya Singla, Sanjana Sanjeev, Jatin Kumar, Adarsh Raj Shivam, Rajiv Ratn Shah | 2024-04-19T08:45:42Z | cs.CL, cs.AI | The rapid progress in the field of natural language processing (NLP) systems
and the expansion of large language models (LLMs) have opened up numerous
opportunities in the field of education and instructional methods. These
advancements offer the potential for tailored learning experiences and
immediate feedback, all delivered through accessible and cost-effective
services. One notable application area for this technological advancement is in
the realm of solving mathematical problems. Mathematical problem-solving not
only requires the ability to decipher complex problem statements but also the
skill to perform precise arithmetic calculations at each step of the
problem-solving process. However, the evaluation of the arithmetic capabilities
of large language models remains an area that has received relatively little
attention. In response, we introduce an extensive mathematics dataset called
"MathQuest" sourced from the 11th and 12th standard Mathematics NCERT
textbooks. This dataset encompasses mathematical challenges of varying
complexity and covers a wide range of mathematical concepts. Utilizing this
dataset, we conduct fine-tuning experiments with three prominent LLMs: LLaMA-2,
WizardMath, and MAmmoTH. These fine-tuned models serve as benchmarks for
evaluating their performance on our dataset. Our experiments reveal that among
the three models, MAmmoTH-13B emerges as the most proficient, achieving the
highest level of competence in solving the presented mathematical problems.
Consequently, MAmmoTH-13B establishes itself as a robust and dependable
benchmark for addressing NCERT mathematics problems. | cs.CL | Mathify: Evaluating Large Language Models on Mathematical Problem Solving Tasks |
2024-04-19T08:06:01Z | http://arxiv.org/pdf/2404.12698v1 | http://arxiv.org/abs/2404.12698v1 | Xiao Zhang, Gosse Bouma, Johan Bos | 2024-04-19T08:06:01Z | cs.CL | Current open-domain neural semantics parsers show impressive performance.
However, closer inspection of the symbolic meaning representations they produce
reveals significant weaknesses: sometimes they tend to merely copy character
sequences from the source text to form symbolic concepts, defaulting to the
most frequent word sense based in the training distribution. By leveraging the
hierarchical structure of a lexical ontology, we introduce a novel
compositional symbolic representation for concepts based on their position in
the taxonomical hierarchy. This representation provides richer semantic
information and enhances interpretability. We introduce a neural "taxonomical"
semantic parser to utilize this new representation system of predicates, and
compare it with a standard neural semantic parser trained on the traditional
meaning representation format, employing a novel challenge set and evaluation
metric for evaluation. Our experimental findings demonstrate that the
taxonomical model, trained on much richer and complex meaning representations,
is slightly subordinate in performance to the traditional model using the
standard metrics for evaluation, but outperforms it when dealing with
out-of-vocabulary concepts. This finding is encouraging for research in
computational semantics that aims to combine data-driven distributional
meanings with knowledge-based symbolic representations. | cs.CL | Neural Semantic Parsing with Extremely Rich Symbolic Meaning Representations |
2024-04-19T07:14:31Z | http://arxiv.org/pdf/2404.12670v1 | http://arxiv.org/abs/2404.12670v1 | Yang Deng, Lizi Liao, Zhonghua Zheng, Grace Hui Yang, Tat-Seng Chua | 2024-04-19T07:14:31Z | cs.IR, cs.CL, cs.HC | Recent research on proactive conversational agents (PCAs) mainly focuses on
improving the system's capabilities in anticipating and planning action
sequences to accomplish tasks and achieve goals before users articulate their
requests. This perspectives paper highlights the importance of moving towards
building human-centered PCAs that emphasize human needs and expectations, and
that considers ethical and social implications of these agents, rather than
solely focusing on technological capabilities. The distinction between a
proactive and a reactive system lies in the proactive system's
initiative-taking nature. Without thoughtful design, proactive systems risk
being perceived as intrusive by human users. We address the issue by
establishing a new taxonomy concerning three key dimensions of human-centered
PCAs, namely Intelligence, Adaptivity, and Civility. We discuss potential
research opportunities and challenges based on this new taxonomy upon the five
stages of PCA system construction. This perspectives paper lays a foundation
for the emerging area of conversational information retrieval research and
paves the way towards advancing human-centered proactive conversational
systems. | cs.IR | Towards Human-centered Proactive Conversational Agents |
2024-04-19T06:58:51Z | http://arxiv.org/pdf/2404.12659v1 | http://arxiv.org/abs/2404.12659v1 | Hongzhi Qi, Hanfei Liu, Jianqiang Li, Qing Zhao, Wei Zhai, Dan Luo, Tian Yu He, Shuo Liu, Bing Xiang Yang, Guanghui Fu | 2024-04-19T06:58:51Z | cs.CL | In the social media, users frequently express personal emotions, a subset of
which may indicate potential suicidal tendencies. The implicit and varied forms
of expression in internet language complicate accurate and rapid identification
of suicidal intent on social media, thus creating challenges for timely
intervention efforts. The development of deep learning models for suicide risk
detection is a promising solution, but there is a notable lack of relevant
datasets, especially in the Chinese context. To address this gap, this study
presents a Chinese social media dataset designed for fine-grained suicide risk
classification, focusing on indicators such as expressions of suicide intent,
methods of suicide, and urgency of timing. Seven pre-trained models were
evaluated in two tasks: high and low suicide risk, and fine-grained suicide
risk classification on a level of 0 to 10. In our experiments, deep learning
models show good performance in distinguishing between high and low suicide
risk, with the best model achieving an F1 score of 88.39%. However, the results
for fine-grained suicide risk classification were still unsatisfactory, with an
weighted F1 score of 50.89%. To address the issues of data imbalance and
limited dataset size, we investigated both traditional and advanced, large
language model based data augmentation techniques, demonstrating that data
augmentation can enhance model performance by up to 4.65% points in F1-score.
Notably, the Chinese MentalBERT model, which was pre-trained on psychological
domain data, shows superior performance in both tasks. This study provides
valuable insights for automatic identification of suicidal individuals,
facilitating timely psychological intervention on social media platforms. The
source code and data are publicly available. | cs.CL | SOS-1K: A Fine-grained Suicide Risk Classification Dataset for Chinese Social Media Analysis |
2024-04-19T06:41:32Z | http://arxiv.org/pdf/2404.12652v1 | http://arxiv.org/abs/2404.12652v1 | Yuan Zang, Tian Yun, Hao Tan, Trung Bui, Chen Sun | 2024-04-19T06:41:32Z | cs.CV, cs.AI, cs.CL, cs.LG | Do vision-language models (VLMs) pre-trained to caption an image of a
"durian" learn visual concepts such as "brown" (color) and "spiky" (texture) at
the same time? We aim to answer this question as visual concepts learned "for
free" would enable wide applications such as neuro-symbolic reasoning or
human-interpretable object classification. We assume that the visual concepts,
if captured by pre-trained VLMs, can be extracted by their vision-language
interface with text-based concept prompts. We observe that recent works
prompting VLMs with concepts often differ in their strategies to define and
evaluate the visual concepts, leading to conflicting conclusions. We propose a
new concept definition strategy based on two observations: First, certain
concept prompts include shortcuts that recognize correct concepts for wrong
reasons; Second, multimodal information (e.g. visual discriminativeness, and
textual knowledge) should be leveraged when selecting the concepts. Our
proposed concept discovery and learning (CDL) framework is thus designed to
identify a diverse list of generic visual concepts (e.g. "spiky" as opposed to
"spiky durian"), which are ranked and selected based on visual and language
mutual information. We carefully design quantitative and human evaluations of
the discovered concepts on six diverse visual recognition datasets, which
confirm that pre-trained VLMs do learn visual concepts that provide accurate
and thorough descriptions for the recognized objects. All code and models are
publicly released. | cs.CV | Pre-trained Vision-Language Models Learn Discoverable Visual Concepts |
2024-04-19T05:48:09Z | http://arxiv.org/pdf/2404.12642v1 | http://arxiv.org/abs/2404.12642v1 | Shanmin Wang, Hui Shuai, Qingshan Liu, Fei Wang | 2024-04-19T05:48:09Z | cs.CL, cs.CV | In this paper, we propose a new Multimodal Representation Learning (MRL)
method for Multimodal Sentiment Analysis (MSA), which facilitates the adaptive
interaction between modalities through Cooperative Sentiment Agents, named
Co-SA. Co-SA comprises two critical components: the Sentiment Agents
Establishment (SAE) phase and the Sentiment Agents Cooperation (SAC) phase.
During the SAE phase, each sentiment agent deals with an unimodal signal and
highlights explicit dynamic sentiment variations within the modality via the
Modality-Sentiment Disentanglement (MSD) and Deep Phase Space Reconstruction
(DPSR) modules. Subsequently, in the SAC phase, Co-SA meticulously designs
task-specific interaction mechanisms for sentiment agents so that coordinating
multimodal signals to learn the joint representation. Specifically, Co-SA
equips an independent policy model for each sentiment agent that captures
significant properties within the modality. These policies are optimized
mutually through the unified reward adaptive to downstream tasks. Benefitting
from the rewarding mechanism, Co-SA transcends the limitation of pre-defined
fusion modes and adaptively captures unimodal properties for MRL in the
multimodal interaction setting. To demonstrate the effectiveness of Co-SA, we
apply it to address Multimodal Sentiment Analysis (MSA) and Multimodal Emotion
Recognition (MER) tasks. Our comprehensive experimental results demonstrate
that Co-SA excels at discovering diverse cross-modal features, encompassing
both common and complementary aspects. The code can be available at
https://github.com/smwanghhh/Co-SA. | cs.CL | Cooperative Sentiment Agents for Multimodal Sentiment Analysis |
2024-04-19T05:01:12Z | http://arxiv.org/pdf/2404.12628v1 | http://arxiv.org/abs/2404.12628v1 | Darshan Prabhu, Sai Ganesh Mirishkar, Pankaj Wasnik | 2024-04-19T05:01:12Z | cs.CL | Self-supervised learned (SSL) models such as Wav2vec and HuBERT yield
state-of-the-art results on speech-related tasks. Given the effectiveness of
such models, it is advantageous to use them in conventional ASR systems. While
some approaches suggest incorporating these models as a trainable encoder or a
learnable frontend, training such systems is extremely slow and requires a lot
of computation cycles. In this work, we propose two simple approaches that use
(1) framewise addition and (2) cross-attention mechanisms to efficiently
incorporate the representations from the SSL model(s) into the ASR
architecture, resulting in models that are comparable in size with standard
encoder-decoder conformer systems while also avoiding the usage of SSL models
during training. Our approach results in faster training and yields significant
performance gains on the Librispeech and Tedlium datasets compared to
baselines. We further provide detailed analysis and ablation studies that
demonstrate the effectiveness of our approach. | cs.CL | Efficient infusion of self-supervised representations in Automatic Speech Recognition |
2024-04-19T04:02:50Z | http://arxiv.org/pdf/2404.12618v1 | http://arxiv.org/abs/2404.12618v1 | Hoang H. Nguyen, Chenwei Zhang, Ye Liu, Natalie Parde, Eugene Rohrbaugh, Philip S. Yu | 2024-04-19T04:02:50Z | cs.CL, cs.AI, cs.LG | Naively assuming English as a source language may hinder cross-lingual
transfer for many languages by failing to consider the importance of language
contact. Some languages are more well-connected than others, and target
languages can benefit from transferring from closely related languages; for
many languages, the set of closely related languages does not include English.
In this work, we study the impact of source language for cross-lingual
transfer, demonstrating the importance of selecting source languages that have
high contact with the target language. We also construct a novel benchmark
dataset for close contact Chinese-Japanese-Korean-Vietnamese (CJKV) languages
to further encourage in-depth studies of language contact. To comprehensively
capture contact between these languages, we propose to integrate Romanized
transcription beyond textual scripts via Contrastive Learning objectives,
leading to enhanced cross-lingual representations and effective zero-shot
cross-lingual transfer. | cs.CL | CORI: CJKV Benchmark with Romanization Integration -- A step towards Cross-lingual Transfer Beyond Textual Scripts |
2024-04-19T03:28:18Z | http://arxiv.org/pdf/2404.12608v1 | http://arxiv.org/abs/2404.12608v1 | Sibei Chen, Yeye He, Weiwei Cui, Ju Fan, Song Ge, Haidong Zhang, Dongmei Zhang, Surajit Chaudhuri | 2024-04-19T03:28:18Z | cs.DB, cs.CL, cs.PL | Spreadsheets are widely recognized as the most popular end-user programming
tools, which blend the power of formula-based computation, with an intuitive
table-based interface. Today, spreadsheets are used by billions of users to
manipulate tables, most of whom are neither database experts nor professional
programmers.
Despite the success of spreadsheets, authoring complex formulas remains
challenging, as non-technical users need to look up and understand non-trivial
formula syntax. To address this pain point, we leverage the observation that
there is often an abundance of similar-looking spreadsheets in the same
organization, which not only have similar data, but also share similar
computation logic encoded as formulas. We develop an Auto-Formula system that
can accurately predict formulas that users want to author in a target
spreadsheet cell, by learning and adapting formulas that already exist in
similar spreadsheets, using contrastive-learning techniques inspired by
"similar-face recognition" from compute vision.
Extensive evaluations on over 2K test formulas extracted from real enterprise
spreadsheets show the effectiveness of Auto-Formula over alternatives. Our
benchmark data is available at https://github.com/microsoft/Auto-Formula to
facilitate future research. | cs.DB | Auto-Formula: Recommend Formulas in Spreadsheets using Contrastive Learning for Table Representations |
2024-04-19T02:59:09Z | http://arxiv.org/pdf/2404.12596v1 | http://arxiv.org/abs/2404.12596v1 | Lasal Jayawardena, Prasan Yapa | 2024-04-19T02:59:09Z | cs.CL, cs.AI, cs.LG | Over the past year, the field of Natural Language Generation (NLG) has
experienced an exponential surge, largely due to the introduction of Large
Language Models (LLMs). These models have exhibited the most effective
performance in a range of domains within the Natural Language Processing and
Generation domains. However, their application in domain-specific tasks, such
as paraphrasing, presents significant challenges. The extensive number of
parameters makes them difficult to operate on commercial hardware, and they
require substantial time for inference, leading to high costs in a production
setting. In this study, we tackle these obstacles by employing LLMs to develop
three distinct models for the paraphrasing field, applying a method referred to
as sequence-level knowledge distillation. These distilled models are capable of
maintaining the quality of paraphrases generated by the LLM. They demonstrate
faster inference times and the ability to generate diverse paraphrases of
comparable quality. A notable characteristic of these models is their ability
to exhibit syntactic diversity while also preserving lexical diversity,
features previously uncommon due to existing data quality issues in datasets
and not typically observed in neural-based approaches. Human evaluation of our
models shows that there is only a 4% drop in performance compared to the LLM
teacher model used in the distillation process, despite being 1000 times
smaller. This research provides a significant contribution to the NLG field,
offering a more efficient and cost-effective solution for paraphrasing tasks. | cs.CL | Parameter Efficient Diverse Paraphrase Generation Using Sequence-Level Knowledge Distillation |
2024-04-19T02:11:41Z | http://arxiv.org/pdf/2404.12580v1 | http://arxiv.org/abs/2404.12580v1 | Anirudh Sundar, Christopher Richardson, William Gay, Larry Heck | 2024-04-19T02:11:41Z | cs.CL, cs.AI, cs.LG | This paper introduces Interactive Tables (iTBLS), a dataset of interactive
conversations situated in tables from scientific articles. This dataset is
designed to facilitate human-AI collaborative problem-solving through
AI-powered multi-task tabular capabilities. In contrast to prior work that
models interactions as factoid QA or procedure synthesis, iTBLS broadens the
scope of interactions to include mathematical reasoning, natural language
manipulation, and expansion of existing tables from natural language
conversation by delineating interactions into one of three tasks:
interpretation, modification, or generation. Additionally, the paper presents a
suite of baseline approaches to iTBLS, utilizing zero-shot prompting and
parameter-efficient fine-tuning for different computing situations. We also
introduce a novel multi-step approach and show how it can be leveraged in
conjunction with parameter-efficient fine-tuning to achieve the
state-of-the-art on iTBLS; outperforming standard parameter-efficient
fine-tuning by up to 15% on interpretation, 18% on modification, and 38% on
generation. | cs.CL | iTBLS: A Dataset of Interactive Conversations Over Tabular Information |
2024-04-19T00:48:30Z | http://arxiv.org/pdf/2404.12560v1 | http://arxiv.org/abs/2404.12560v1 | Dayton G. Thorpe, Andrew J. Duberstein, Ian A. Kinsey | 2024-04-19T00:48:30Z | cs.CL, cs.DB | The current state-of-the-art (SOTA) for automated text-to-SQL still falls
well short of expert human performance as measured by execution accuracy (EX)
on the BIRD-SQL benchmark. The most accurate methods are also slow and
expensive. To advance the SOTA for text-to-SQL while reducing cost and
improving speed, we explore the combination of low-cost fine tuning, novel
methods for diverse retrieval-augmented generation (RAG) and new input and
output formats that help large language models (LLMs) achieve higher EX. We
introduce two new methods, Dubo-SQL v1 and v2. Dubo-SQL v1 sets a new record
for EX on the holdout test set of BIRD-SQL. Dubo-SQL v2 achieves even higher
performance on the BIRD-SQL dev set. Dubo-SQL v1 relies on LLMs from OpenAI,
but uses the low-cost GPT-3.5 Turbo while exceeding the performance of the
next-best model using OpenAI, which instead uses the more expensive GPT-4.
Dubo-SQL v1 exceeds the performance of the next-best model using GPT-3.5 by
over 20%. Dubo-SQL v2 uses GPT-4 Turbo and RAG in place of fine tuning to push
EX higher. | cs.CL | Dubo-SQL: Diverse Retrieval-Augmented Generation and Fine Tuning for Text-to-SQL |
2024-04-18T23:50:50Z | http://arxiv.org/pdf/2404.12545v2 | http://arxiv.org/abs/2404.12545v2 | Xuemin Yu, Fahim Dalvi, Nadir Durrani, Marzia Nouri, Hassan Sajjad | 2024-06-17T20:24:49Z | cs.CL | Interpreting and understanding the predictions made by deep learning models
poses a formidable challenge due to their inherently opaque nature. Many
previous efforts aimed at explaining these predictions rely on input features,
specifically, the words within NLP models. However, such explanations are often
less informative due to the discrete nature of these words and their lack of
contextual verbosity. To address this limitation, we introduce the Latent
Concept Attribution method (LACOAT), which generates explanations for
predictions based on latent concepts. Our foundational intuition is that a word
can exhibit multiple facets, contingent upon the context in which it is used.
Therefore, given a word in context, the latent space derived from our training
process reflects a specific facet of that word. LACOAT functions by mapping the
representations of salient input words into the training latent space, allowing
it to provide latent context-based explanations of the prediction. | cs.CL | Latent Concept-based Explanation of NLP Models |
2024-04-18T22:56:57Z | http://arxiv.org/pdf/2404.12535v1 | http://arxiv.org/abs/2404.12535v1 | William Watson, Nicole Cho | 2024-04-18T22:56:57Z | cs.LG, cs.AI, cs.CL | Hallucination continues to be one of the most critical challenges in the
institutional adoption journey of Large Language Models (LLMs). In this
context, an overwhelming number of studies have focused on analyzing the
post-generation phase - refining outputs via feedback, analyzing logit output
values, or deriving clues via the outputs' artifacts. We propose HalluciBot, a
model that predicts the probability of hallucination $\textbf{before
generation}$, for any query imposed to an LLM. In essence, HalluciBot does not
invoke any generation during inference. To derive empirical evidence for
HalluciBot, we employ a Multi-Agent Monte Carlo Simulation using a Query
Perturbator to craft $n$ variations per query at train time. The construction
of our Query Perturbator is motivated by our introduction of a new definition
of hallucination - $\textit{truthful hallucination}$. Our training methodology
generated 2,219,022 estimates for a training corpus of 369,837 queries,
spanning 13 diverse datasets and 3 question-answering scenarios. HalluciBot
predicts both binary and multi-class probabilities of hallucination, enabling a
means to judge the query's quality with regards to its propensity to
hallucinate. Therefore, HalluciBot paves the way to revise or cancel a query
before generation and the ensuing computational waste. Moreover, it provides a
lucid means to measure user accountability for hallucinatory queries. | cs.LG | HalluciBot: Is There No Such Thing as a Bad Question? |
2024-04-18T22:01:56Z | http://arxiv.org/pdf/2404.12526v1 | http://arxiv.org/abs/2404.12526v1 | James Seale Smith, Lazar Valkov, Shaunak Halbe, Vyshnavi Gutta, Rogerio Feris, Zsolt Kira, Leonid Karlinsky | 2024-04-18T22:01:56Z | cs.LG, cs.CL, cs.CV | Foundation Models (FMs) have become the hallmark of modern AI, however, these
models are trained on massive data, leading to financially expensive training.
Updating FMs as new data becomes available is important, however, can lead to
`catastrophic forgetting', where models underperform on tasks related to data
sub-populations observed too long ago. This continual learning (CL) phenomenon
has been extensively studied, but primarily in a setting where only a small
amount of past data can be stored. We advocate for the paradigm where memory is
abundant, allowing us to keep all previous data, but computational resources
are limited. In this setting, traditional replay-based CL approaches are
outperformed by a simple baseline which replays past data selected uniformly at
random, indicating that this setting necessitates a new approach. We address
this by introducing a framework of adaptive memory replay for continual
learning, where sampling of past data is phrased as a multi-armed bandit
problem. We utilize Bolzmann sampling to derive a method which dynamically
selects past data for training conditioned on the current task, assuming full
data access and emphasizing training efficiency. Through extensive evaluations
on both vision and language pre-training tasks, we demonstrate the
effectiveness of our approach, which maintains high performance while reducing
forgetting by up to 10% at no training efficiency cost. | cs.LG | Adaptive Memory Replay for Continual Learning |
2024-04-18T20:43:08Z | http://arxiv.org/pdf/2404.12500v1 | http://arxiv.org/abs/2404.12500v1 | Jason Wu, Yi-Hao Peng, Amanda Li, Amanda Swearngin, Jeffrey P. Bigham, Jeffrey Nichols | 2024-04-18T20:43:08Z | cs.HC, cs.CL, cs.CV | User interface (UI) design is a difficult yet important task for ensuring the
usability, accessibility, and aesthetic qualities of applications. In our
paper, we develop a machine-learned model, UIClip, for assessing the design
quality and visual relevance of a UI given its screenshot and natural language
description. To train UIClip, we used a combination of automated crawling,
synthetic augmentation, and human ratings to construct a large-scale dataset of
UIs, collated by description and ranked by design quality. Through training on
the dataset, UIClip implicitly learns properties of good and bad designs by i)
assigning a numerical score that represents a UI design's relevance and quality
and ii) providing design suggestions. In an evaluation that compared the
outputs of UIClip and other baselines to UIs rated by 12 human designers, we
found that UIClip achieved the highest agreement with ground-truth rankings.
Finally, we present three example applications that demonstrate how UIClip can
facilitate downstream applications that rely on instantaneous assessment of UI
design quality: i) UI code generation, ii) UI design tips generation, and iii)
quality-aware UI example search. | cs.HC | UIClip: A Data-driven Model for Assessing User Interface Design |
2024-04-18T20:17:23Z | http://arxiv.org/pdf/2404.12494v1 | http://arxiv.org/abs/2404.12494v1 | Yu Feng, Ben Zhou, Weidong Lin, Dan Roth | 2024-04-18T20:17:23Z | cs.CL | Large language models primarily rely on inductive reasoning for decision
making. This results in unreliable decisions when applied to real-world tasks
that often present incomplete contexts and conditions. Thus, accurate
probability estimation and appropriate interpretations are required to enhance
decision-making reliability. In this paper, we propose a Bayesian inference
framework called BIRD for large language models. BIRD provides controllable and
interpretable probability estimation for model decisions, based on abductive
factors, LLM entailment, as well as learnable deductive Bayesian modeling.
Experiments show that BIRD produces probability estimations that align with
human judgments over 65% of the time using open-sourced Llama models,
outperforming the state-of-the-art GPT-4 by 35%. We also show that BIRD can be
directly used for trustworthy decision making on many real-world applications. | cs.CL | BIRD: A Trustworthy Bayesian Inference Framework for Large Language Models |
2024-04-18T20:15:48Z | http://arxiv.org/pdf/2404.12493v1 | http://arxiv.org/abs/2404.12493v1 | Urchade Zaratiana, Nadi Tomeh, Yann Dauxais, Pierre Holat, Thierry Charnois | 2024-04-18T20:15:48Z | cs.CL, cs.AI | Joint entity and relation extraction plays a pivotal role in various
applications, notably in the construction of knowledge graphs. Despite recent
progress, existing approaches often fall short in two key aspects: richness of
representation and coherence in output structure. These models often rely on
handcrafted heuristics for computing entity and relation representations,
potentially leading to loss of crucial information. Furthermore, they disregard
task and/or dataset-specific constraints, resulting in output structures that
lack coherence. In our work, we introduce EnriCo, which mitigates these
shortcomings. Firstly, to foster rich and expressive representation, our model
leverage attention mechanisms that allow both entities and relations to
dynamically determine the pertinent information required for accurate
extraction. Secondly, we introduce a series of decoding algorithms designed to
infer the highest scoring solutions while adhering to task and dataset-specific
constraints, thus promoting structured and coherent outputs. Our model
demonstrates competitive performance compared to baselines when evaluated on
Joint IE datasets. | cs.CL | EnriCo: Enriched Representation and Globally Constrained Inference for Entity and Relation Extraction |
2024-04-18T20:09:37Z | http://arxiv.org/pdf/2404.12491v1 | http://arxiv.org/abs/2404.12491v1 | Urchade Zaratiana, Nadi Tomeh, Niama El Khbir, Pierre Holat, Thierry Charnois | 2024-04-18T20:09:37Z | cs.CL, cs.AI | Information extraction (IE) is an important task in Natural Language
Processing (NLP), involving the extraction of named entities and their
relationships from unstructured text. In this paper, we propose a novel
approach to this task by formulating it as graph structure learning (GSL). By
formulating IE as GSL, we enhance the model's ability to dynamically refine and
optimize the graph structure during the extraction process. This formulation
allows for better interaction and structure-informed decisions for entity and
relation prediction, in contrast to previous models that have separate or
untied predictions for these tasks. When compared against state-of-the-art
baselines on joint entity and relation extraction benchmarks, our model,
GraphER, achieves competitive results. | cs.CL | GraphER: A Structure-aware Text-to-Graph Model for Entity and Relation Extraction |
2024-04-18T20:05:30Z | http://arxiv.org/pdf/2404.12489v2 | http://arxiv.org/abs/2404.12489v2 | Kelvin Wey Han Chan, Christopher Bryant, Li Nguyen, Andrew Caines, Zheng Yuan | 2024-05-06T22:27:36Z | cs.CL | Code-switching (CSW) is a common phenomenon among multilingual speakers where
multiple languages are used in a single discourse or utterance. Mixed language
utterances may still contain grammatical errors however, yet most existing
Grammar Error Correction (GEC) systems have been trained on monolingual data
and not developed with CSW in mind. In this work, we conduct the first
exploration into the use of GEC systems on CSW text. Through this exploration,
we propose a novel method of generating synthetic CSW GEC datasets by
translating different spans of text within existing GEC corpora. We then
investigate different methods of selecting these spans based on CSW ratio,
switch-point factor and linguistic constraints, and identify how they affect
the performance of GEC systems on CSW text. Our best model achieves an average
increase of 1.57 $F_{0.5}$ across 3 CSW test sets (English-Chinese,
English-Korean and English-Japanese) without affecting the model's performance
on a monolingual dataset. We furthermore discovered that models trained on one
CSW language generalise relatively well to other typologically similar CSW
languages. | cs.CL | Grammatical Error Correction for Code-Switched Sentences by Learners of English |
2024-04-18T19:41:05Z | http://arxiv.org/pdf/2404.14432v1 | http://arxiv.org/abs/2404.14432v1 | Abdul Wahab Ziaullah, Ferda Ofli, Muhammad Imran | 2024-04-18T19:41:05Z | cs.SI, cs.AI, cs.CL, cs.IR | Critical Infrastructure Facilities (CIFs), such as healthcare and
transportation facilities, are vital for the functioning of a community,
especially during large-scale emergencies. In this paper, we explore a
potential application of Large Language Models (LLMs) to monitor the status of
CIFs affected by natural disasters through information disseminated in social
media networks. To this end, we analyze social media data from two disaster
events in two different countries to identify reported impacts to CIFs as well
as their impact severity and operational status. We employ state-of-the-art
open-source LLMs to perform computational tasks including retrieval,
classification, and inference, all in a zero-shot setting. Through extensive
experimentation, we report the results of these tasks using standard evaluation
metrics and reveal insights into the strengths and weaknesses of LLMs. We note
that although LLMs perform well in classification tasks, they encounter
challenges with inference tasks, especially when the context/prompt is complex
and lengthy. Additionally, we outline various potential directions for future
exploration that can be beneficial during the initial adoption phase of LLMs
for disaster response tasks. | cs.SI | Monitoring Critical Infrastructure Facilities During Disasters Using Large Language Models |
2024-04-18T18:48:50Z | http://arxiv.org/pdf/2404.12464v4 | http://arxiv.org/abs/2404.12464v4 | Abhinav Rao, Akhila Yerukola, Vishwa Shah, Katharina Reinecke, Maarten Sap | 2024-06-06T16:02:39Z | cs.CL | The integration of Large Language Models (LLMs) into various global cultures
fundamentally presents a cultural challenge: LLMs must navigate interactions,
respect social norms, and avoid transgressing cultural boundaries. However, it
is still unclear if LLMs can adapt their outputs to diverse cultural norms. Our
study focuses on this aspect. We introduce NormAd, a novel dataset, which
includes 2.6k stories that represent social and cultural norms from 75
countries, to assess the ability of LLMs to adapt to different granular levels
of socio-cultural contexts such as the country of origin, its associated
cultural values, and prevalent social norms. Our study reveals that LLMs
struggle with cultural reasoning across all contextual granularities, showing
stronger adaptability to English-centric cultures over those from the Global
South. Even with explicit social norms, the top-performing model,
Mistral-7b-Instruct, achieves only 81.8\% accuracy, lagging behind the 95.6\%
achieved by humans. Evaluation on NormAd further reveals that LLMs struggle to
adapt to stories involving gift-giving across cultures. Due to inherent
agreement or sycophancy biases, LLMs find it considerably easier to assess the
social acceptability of stories that adhere to cultural norms than those that
deviate from them. Our benchmark measures the cultural adaptability (or lack
thereof) of LLMs, emphasizing the potential to make these technologies more
equitable and useful for global audiences. We release the NormAd dataset and
its associated code on GitHub. | cs.CL | NormAd: A Benchmark for Measuring the Cultural Adaptability of Large Language Models |
2024-04-18T18:32:30Z | http://arxiv.org/pdf/2404.12457v2 | http://arxiv.org/abs/2404.12457v2 | Chao Jin, Zili Zhang, Xuanlin Jiang, Fangyue Liu, Xin Liu, Xuanzhe Liu, Xin Jin | 2024-04-25T06:47:57Z | cs.DC, cs.CL, cs.LG | Retrieval-Augmented Generation (RAG) has shown significant improvements in
various natural language processing tasks by integrating the strengths of large
language models (LLMs) and external knowledge databases. However, RAG
introduces long sequence generation and leads to high computation and memory
costs. We propose RAGCache, a novel multilevel dynamic caching system tailored
for RAG. Our analysis benchmarks current RAG systems, pinpointing the
performance bottleneck (i.e., long sequence due to knowledge injection) and
optimization opportunities (i.e., caching knowledge's intermediate states).
Based on these insights, we design RAGCache, which organizes the intermediate
states of retrieved knowledge in a knowledge tree and caches them in the GPU
and host memory hierarchy. RAGCache proposes a replacement policy that is aware
of LLM inference characteristics and RAG retrieval patterns. It also
dynamically overlaps the retrieval and inference steps to minimize the
end-to-end latency. We implement RAGCache and evaluate it on vLLM, a
state-of-the-art LLM inference system and Faiss, a state-of-the-art vector
database. The experimental results show that RAGCache reduces the time to first
token (TTFT) by up to 4x and improves the throughput by up to 2.1x compared to
vLLM integrated with Faiss. | cs.DC | RAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generation |
2024-04-18T18:26:43Z | http://arxiv.org/pdf/2404.12452v1 | http://arxiv.org/abs/2404.12452v1 | Bingbing Wen, Bill Howe, Lucy Lu Wang | 2024-04-18T18:26:43Z | cs.CL | The correct model response in the face of uncertainty is to abstain from
answering a question so as not to mislead the user. In this work, we study the
ability of LLMs to abstain from answering context-dependent science questions
when provided insufficient or incorrect context. We probe model sensitivity in
several settings: removing gold context, replacing gold context with irrelevant
context, and providing additional context beyond what is given. In experiments
on four QA datasets with four LLMs, we show that performance varies greatly
across models, across the type of context provided, and also by question type;
in particular, many LLMs seem unable to abstain from answering boolean
questions using standard QA prompts. Our analysis also highlights the
unexpected impact of abstention performance on QA task accuracy.
Counter-intuitively, in some settings, replacing gold context with irrelevant
context or adding irrelevant context to gold context can improve abstention
performance in a way that results in improvements in task performance. Our
results imply that changes are needed in QA dataset design and evaluation to
more effectively assess the correctness and downstream impacts of model
abstention. | cs.CL | Characterizing LLM Abstention Behavior in Science QA with Context Perturbations |
2024-04-18T18:12:01Z | http://arxiv.org/pdf/2404.12447v2 | http://arxiv.org/abs/2404.12447v2 | Yoonsang Lee, Xi Ye, Eunsol Choi | 2024-05-26T14:00:07Z | cs.CL | Different entities with the same name can be difficult to distinguish.
Handling confusing entity mentions is a crucial skill for language models
(LMs). For example, given the question "Where was Michael Jordan educated?" and
a set of documents discussing different people named Michael Jordan, can LMs
distinguish entity mentions to generate a cohesive answer to the question? To
test this ability, we introduce a new benchmark, AmbigDocs. By leveraging
Wikipedia's disambiguation pages, we identify a set of documents, belonging to
different entities who share an ambiguous name. From these documents, we
generate questions containing an ambiguous name and their corresponding sets of
answers. Our analysis reveals that current state-of-the-art models often yield
ambiguous answers or incorrectly merge information belonging to different
entities. We establish an ontology categorizing four types of incomplete
answers and automatic evaluation metrics to identify such categories. We lay
the foundation for future work on reasoning across multiple documents with
ambiguous entities. | cs.CL | AmbigDocs: Reasoning across Documents on Different Entities under the Same Name |
2024-04-18T18:03:08Z | http://arxiv.org/pdf/2404.12444v1 | http://arxiv.org/abs/2404.12444v1 | Tianze Hua, Tian Yun, Ellie Pavlick | 2024-04-18T18:03:08Z | cs.CL, cs.AI | Many pretrained multilingual models exhibit cross-lingual transfer ability,
which is often attributed to a learned language-neutral representation during
pretraining. However, it remains unclear what factors contribute to the
learning of a language-neutral representation, and whether the learned
language-neutral representation suffices to facilitate cross-lingual transfer.
We propose a synthetic task, Multilingual Othello (mOthello), as a testbed to
delve into these two questions. We find that: (1) models trained with naive
multilingual pretraining fail to learn a language-neutral representation across
all input languages; (2) the introduction of "anchor tokens" (i.e., lexical
items that are identical across languages) helps cross-lingual representation
alignment; and (3) the learning of a language-neutral representation alone is
not sufficient to facilitate cross-lingual transfer. Based on our findings, we
propose a novel approach - multilingual pretraining with unified output space -
that both induces the learning of language-neutral representation and
facilitates cross-lingual transfer. | cs.CL | mOthello: When Do Cross-Lingual Representation Alignment and Cross-Lingual Transfer Emerge in Multilingual Models? |
2024-04-18T17:59:54Z | http://arxiv.org/pdf/2404.12390v4 | http://arxiv.org/abs/2404.12390v4 | Xingyu Fu, Yushi Hu, Bangzheng Li, Yu Feng, Haoyu Wang, Xudong Lin, Dan Roth, Noah A. Smith, Wei-Chiu Ma, Ranjay Krishna | 2024-07-03T08:44:45Z | cs.CV, cs.AI, cs.CL | We introduce Blink, a new benchmark for multimodal language models (LLMs)
that focuses on core visual perception abilities not found in other
evaluations. Most of the Blink tasks can be solved by humans "within a blink"
(e.g., relative depth estimation, visual correspondence, forensics detection,
and multi-view reasoning). However, we find these perception-demanding tasks
cast significant challenges for current multimodal LLMs because they resist
mediation through natural language. Blink reformats 14 classic computer vision
tasks into 3,807 multiple-choice questions, paired with single or multiple
images and visual prompting. While humans get 95.70% accuracy on average, Blink
is surprisingly challenging for existing multimodal LLMs: even the
best-performing GPT-4V and Gemini achieve accuracies of 51.26% and 45.72%, only
13.17% and 7.63% higher than random guessing, indicating that such perception
abilities have not "emerged" yet in recent multimodal LLMs. Our analysis also
highlights that specialist CV models could solve these problems much better,
suggesting potential pathways for future improvements. We believe Blink will
stimulate the community to help multimodal LLMs catch up with human-level
visual perception. | cs.CV | BLINK: Multimodal Large Language Models Can See but Not Perceive |
2024-04-18T17:59:48Z | http://arxiv.org/pdf/2404.12387v1 | http://arxiv.org/abs/2404.12387v1 | Reka Team, Aitor Ormazabal, Che Zheng, Cyprien de Masson d'Autume, Dani Yogatama, Deyu Fu, Donovan Ong, Eric Chen, Eugenie Lamprecht, Hai Pham, Isaac Ong, Kaloyan Aleksiev, Lei Li, Matthew Henderson, Max Bain, Mikel Artetxe, Nishant Relan, Piotr Padlewski, Qi Liu, Ren Chen, Samuel Phua, Yazheng Yang, Yi Tay, Yuqi Wang, Zhongkai Zhu, Zhihui Xie | 2024-04-18T17:59:48Z | cs.CL, cs.CV | We introduce Reka Core, Flash, and Edge, a series of powerful multimodal
language models trained from scratch by Reka. Reka models are able to process
and reason with text, images, video, and audio inputs. This technical report
discusses details of training some of these models and provides comprehensive
evaluation results. We show that Reka Edge and Reka Flash are not only
state-of-the-art but also outperform many much larger models, delivering
outsized values for their respective compute class. Meanwhile, our most capable
and largest model, Reka Core, approaches the best frontier models on both
automatic evaluations and blind human evaluations. On image question answering
benchmarks (e.g. MMMU, VQAv2), Core performs competitively to GPT4-V.
Meanwhile, on multimodal chat, Core ranks as the second most preferred model
under a blind third-party human evaluation setup, outperforming other models
such as Claude 3 Opus. On text benchmarks, Core not only performs competitively
to other frontier models on a set of well-established benchmarks (e.g. MMLU,
GSM8K) but also outperforms GPT4-0613 on human evaluation. On video question
answering (Perception-Test), Core outperforms Gemini Ultra. Models are shipped
in production at http://chat.reka.ai . A showcase of non cherry picked
qualitative examples can also be found at http://showcase.reka.ai . | cs.CL | Reka Core, Flash, and Edge: A Series of Powerful Multimodal Language Models |
2024-04-18T17:48:05Z | http://arxiv.org/pdf/2404.12365v1 | http://arxiv.org/abs/2404.12365v1 | Asaf Yehudai, Elron Bendel | 2024-04-18T17:48:05Z | cs.CL, cs.AI, cs.IR, cs.LG | We present FastFit, a method, and a Python package design to provide fast and
accurate few-shot classification, especially for scenarios with many
semantically similar classes. FastFit utilizes a novel approach integrating
batch contrastive learning and token-level similarity score. Compared to
existing few-shot learning packages, such as SetFit, Transformers, or few-shot
prompting of large language models via API calls, FastFit significantly
improves multiclass classification performance in speed and accuracy across
FewMany, our newly curated English benchmark, and Multilingual datasets.
FastFit demonstrates a 3-20x improvement in training speed, completing training
in just a few seconds. The FastFit package is now available on GitHub and PyPi,
presenting a user-friendly solution for NLP practitioners. | cs.CL | When LLMs are Unfit Use FastFit: Fast and Effective Text Classification with Many Classes |
2024-04-18T17:16:16Z | http://arxiv.org/pdf/2404.12342v1 | http://arxiv.org/abs/2404.12342v1 | Nicolay Rusnachenko, Anton Golubev, Natalia Loukachevitch | 2024-04-18T17:16:16Z | cs.CL | In this paper we investigate the use of decoder-based generative transformers
for extracting sentiment towards the named entities in Russian news articles.
We study sentiment analysis capabilities of instruction-tuned large language
models (LLMs). We consider the dataset of RuSentNE-2023 in our study. The first
group of experiments was aimed at the evaluation of zero-shot capabilities of
LLMs with closed and open transparencies. The second covers the fine-tuning of
Flan-T5 using the "chain-of-thought" (CoT) three-hop reasoning framework
(THoR). We found that the results of the zero-shot approaches are similar to
the results achieved by baseline fine-tuned encoder-based transformers
(BERT-base). Reasoning capabilities of the fine-tuned Flan-T5 models with THoR
achieve at least 5% increment with the base-size model compared to the results
of the zero-shot experiment. The best results of sentiment analysis on
RuSentNE-2023 were achieved by fine-tuned Flan-T5-xl, which surpassed the
results of previous state-of-the-art transformer-based classifiers. Our CoT
application framework is publicly available:
https://github.com/nicolay-r/Reasoning-for-Sentiment-Analysis-Framework | cs.CL | Large Language Models in Targeted Sentiment Analysis |
2024-04-18T16:52:36Z | http://arxiv.org/pdf/2404.12318v1 | http://arxiv.org/abs/2404.12318v1 | Zhaofeng Wu, Ananth Balashankar, Yoon Kim, Jacob Eisenstein, Ahmad Beirami | 2024-04-18T16:52:36Z | cs.CL | Aligning language models (LMs) based on human-annotated preference data is a
crucial step in obtaining practical and performant LM-based systems. However,
multilingual human preference data are difficult to obtain at scale, making it
challenging to extend this framework to diverse languages. In this work, we
evaluate a simple approach for zero-shot cross-lingual alignment, where a
reward model is trained on preference data in one source language and directly
applied to other target languages. On summarization and open-ended dialog
generation, we show that this method is consistently successful under
comprehensive evaluation settings, including human evaluation: cross-lingually
aligned models are preferred by humans over unaligned models on up to >70% of
evaluation instances. We moreover find that a different-language reward model
sometimes yields better aligned models than a same-language reward model. We
also identify best practices when there is no language-specific data for even
supervised finetuning, another component in alignment. | cs.CL | Reuse Your Rewards: Reward Model Transfer for Zero-Shot Cross-Lingual Alignment |
2024-04-18T16:24:12Z | http://arxiv.org/pdf/2404.12299v1 | http://arxiv.org/abs/2404.12299v1 | Yusuke Sakai, Mana Makinae, Hidetaka Kamigaito, Taro Watanabe | 2024-04-18T16:24:12Z | cs.CL, cs.AI, cs.LG, cs.SD, eess.AS | In Simultaneous Machine Translation (SiMT) systems, training with a
simultaneous interpretation (SI) corpus is an effective method for achieving
high-quality yet low-latency systems. However, it is very challenging to curate
such a corpus due to limitations in the abilities of annotators, and hence,
existing SI corpora are limited. Therefore, we propose a method to convert
existing speech translation corpora into interpretation-style data, maintaining
the original word order and preserving the entire source content using Large
Language Models (LLM-SI-Corpus). We demonstrate that fine-tuning SiMT models in
text-to-text and speech-to-text settings with the LLM-SI-Corpus reduces
latencies while maintaining the same level of quality as the models trained
with offline datasets. The LLM-SI-Corpus is available at
\url{https://github.com/yusuke1997/LLM-SI-Corpus}. | cs.CL | Simultaneous Interpretation Corpus Construction by Large Language Models in Distant Language Pair |
2024-04-18T16:11:17Z | http://arxiv.org/pdf/2404.12291v2 | http://arxiv.org/abs/2404.12291v2 | Yucheng Lin, Yuhan Xia, Yunfei Long | 2024-04-20T01:52:29Z | cs.CL, cs.AI | This study introduces a novel method for irony detection, applying Large
Language Models (LLMs) with prompt-based learning to facilitate emotion-centric
text augmentation. Traditional irony detection techniques typically fall short
due to their reliance on static linguistic features and predefined knowledge
bases, often overlooking the nuanced emotional dimensions integral to irony. In
contrast, our methodology augments the detection process by integrating subtle
emotional cues, augmented through LLMs, into three benchmark pre-trained NLP
models - BERT, T5, and GPT-2 - which are widely recognized as foundational in
irony detection. We assessed our method using the SemEval-2018 Task 3 dataset
and observed substantial enhancements in irony detection capabilities. | cs.CL | Augmenting emotion features in irony detection with Large language modeling |
2024-04-18T16:10:38Z | http://arxiv.org/pdf/2404.12289v1 | http://arxiv.org/abs/2404.12289v1 | Simeon Junker, Sina Zarrieß | 2024-04-18T16:10:38Z | cs.CL | Scene context is well known to facilitate humans' perception of visible
objects. In this paper, we investigate the role of context in Referring
Expression Generation (REG) for objects in images, where existing research has
often focused on distractor contexts that exert pressure on the generator. We
take a new perspective on scene context in REG and hypothesize that contextual
information can be conceived of as a resource that makes REG models more
resilient and facilitates the generation of object descriptions, and object
types in particular. We train and test Transformer-based REG models with target
representations that have been artificially obscured with noise to varying
degrees. We evaluate how properties of the models' visual context affect their
processing and performance. Our results show that even simple scene contexts
make models surprisingly resilient to perturbations, to the extent that they
can identify referent types even when visual information about the target is
completely missing. | cs.CL | Resilience through Scene Context in Visual Referring Expression Generation |
2024-04-18T15:58:56Z | http://arxiv.org/pdf/2404.12283v1 | http://arxiv.org/abs/2404.12283v1 | Nicholas Harris, Anand Butani, Syed Hashmy | 2024-04-18T15:58:56Z | cs.CL | Embedding models are crucial for various natural language processing tasks
but can be limited by factors such as limited vocabulary, lack of context, and
grammatical errors. This paper proposes a novel approach to improve embedding
performance by leveraging large language models (LLMs) to enrich and rewrite
input text before the embedding process. By utilizing ChatGPT 3.5 to provide
additional context, correct inaccuracies, and incorporate metadata, the
proposed method aims to enhance the utility and accuracy of embedding models.
The effectiveness of this approach is evaluated on three datasets:
Banking77Classification, TwitterSemEval 2015, and Amazon Counter-factual
Classification. Results demonstrate significant improvements over the baseline
model on the TwitterSemEval 2015 dataset, with the best-performing prompt
achieving a score of 85.34 compared to the previous best of 81.52 on the
Massive Text Embedding Benchmark (MTEB) Leaderboard. However, performance on
the other two datasets was less impressive, highlighting the importance of
considering domain-specific characteristics. The findings suggest that
LLM-based text enrichment has shown promising results to improve embedding
performance, particularly in certain domains. Hence, numerous limitations in
the process of embedding can be avoided. | cs.CL | Enhancing Embedding Performance through Large Language Model-based Text Enrichment and Rewriting |
2024-04-18T15:47:00Z | http://arxiv.org/pdf/2404.12274v1 | http://arxiv.org/abs/2404.12274v1 | Jiabao Ji, Bairu Hou, Zhen Zhang, Guanhua Zhang, Wenqi Fan, Qing Li, Yang Zhang, Gaowen Liu, Sijia Liu, Shiyu Chang | 2024-04-18T15:47:00Z | cs.CL, cs.AI | Although large language models (LLMs) have achieved significant success,
their vulnerability to adversarial perturbations, including recent jailbreak
attacks, has raised considerable concerns. However, the increasing size of
these models and their limited access make improving their robustness a
challenging task. Among various defense strategies, randomized smoothing has
shown great potential for LLMs, as it does not require full access to the
model's parameters or fine-tuning via adversarial training. However, randomized
smoothing involves adding noise to the input before model prediction, and the
final model's robustness largely depends on the model's performance on these
noise corrupted data. Its effectiveness is often limited by the model's
sub-optimal performance on noisy data. To address this issue, we propose to
leverage the multitasking nature of LLMs to first denoise the noisy inputs and
then to make predictions based on these denoised versions. We call this
procedure self-denoised smoothing. Unlike previous denoised smoothing
techniques in computer vision, which require training a separate model to
enhance the robustness of LLMs, our method offers significantly better
efficiency and flexibility. Our experimental results indicate that our method
surpasses existing methods in both empirical and certified robustness in
defending against adversarial attacks for both downstream tasks and human
alignments (i.e., jailbreak attacks). Our code is publicly available at
https://github.com/UCSB-NLP-Chang/SelfDenoise | cs.CL | Advancing the Robustness of Large Language Models through Self-Denoised Smoothing |
2024-04-18T15:46:26Z | http://arxiv.org/pdf/2404.12273v1 | http://arxiv.org/abs/2404.12273v1 | Yuanqin He, Yan Kang, Lixin Fan, Qiang Yang | 2024-04-18T15:46:26Z | cs.AI, cs.CL, cs.LG | Federated Learning (FL) has emerged as a promising solution for collaborative
training of large language models (LLMs). However, the integration of LLMs into
FL introduces new challenges, particularly concerning the evaluation of LLMs.
Traditional evaluation methods that rely on labeled test sets and
similarity-based metrics cover only a subset of the acceptable answers, thereby
failing to accurately reflect the performance of LLMs on generative tasks.
Meanwhile, although automatic evaluation methods that leverage advanced LLMs
present potential, they face critical risks of data leakage due to the need to
transmit data to external servers and suboptimal performance on downstream
tasks due to the lack of domain knowledge. To address these issues, we propose
a Federated Evaluation framework of Large Language Models, named FedEval-LLM,
that provides reliable performance measurements of LLMs on downstream tasks
without the reliance on labeled test sets and external tools, thus ensuring
strong privacy-preserving capability. FedEval-LLM leverages a consortium of
personalized LLMs from participants as referees to provide domain knowledge and
collective evaluation capability, thus aligning to the respective downstream
tasks and mitigating uncertainties and biases associated with a single referee.
Experimental results demonstrate a significant improvement in the evaluation
capability of personalized evaluation models on downstream tasks. When applied
to FL, these evaluation models exhibit strong agreement with human preference
and RougeL-score on meticulously curated test sets. FedEval-LLM effectively
overcomes the limitations of traditional metrics and the reliance on external
services, making it a promising framework for the evaluation of LLMs within
collaborative training scenarios. | cs.AI | FedEval-LLM: Federated Evaluation of Large Language Models on Downstream Tasks with Collective Wisdom |
2024-04-18T15:21:34Z | http://arxiv.org/pdf/2404.12253v1 | http://arxiv.org/abs/2404.12253v1 | Ye Tian, Baolin Peng, Linfeng Song, Lifeng Jin, Dian Yu, Haitao Mi, Dong Yu | 2024-04-18T15:21:34Z | cs.CL, cs.LG | Despite the impressive capabilities of Large Language Models (LLMs) on
various tasks, they still struggle with scenarios that involves complex
reasoning and planning. Recent work proposed advanced prompting techniques and
the necessity of fine-tuning with high-quality data to augment LLMs' reasoning
abilities. However, these approaches are inherently constrained by data
availability and quality. In light of this, self-correction and self-learning
emerge as viable solutions, employing strategies that allow LLMs to refine
their outputs and learn from self-assessed rewards. Yet, the efficacy of LLMs
in self-refining its response, particularly in complex reasoning and planning
task, remains dubious. In this paper, we introduce AlphaLLM for the
self-improvements of LLMs, which integrates Monte Carlo Tree Search (MCTS) with
LLMs to establish a self-improving loop, thereby enhancing the capabilities of
LLMs without additional annotations. Drawing inspiration from the success of
AlphaGo, AlphaLLM addresses the unique challenges of combining MCTS with LLM
for self-improvement, including data scarcity, the vastness search spaces of
language tasks, and the subjective nature of feedback in language tasks.
AlphaLLM is comprised of prompt synthesis component, an efficient MCTS approach
tailored for language tasks, and a trio of critic models for precise feedback.
Our experimental results in mathematical reasoning tasks demonstrate that
AlphaLLM significantly enhances the performance of LLMs without additional
annotations, showing the potential for self-improvement in LLMs. | cs.CL | Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing |
2024-04-18T15:02:35Z | http://arxiv.org/pdf/2404.12242v1 | http://arxiv.org/abs/2404.12242v1 | Mengna Zhu, Zijie Xu, Kaisheng Zeng, Kaiming Xiao, Mao Wang, Wenjun Ke, Hongbin Huang | 2024-04-18T15:02:35Z | cs.CL | Extracting structured event knowledge, including event triggers and
corresponding arguments, from military texts is fundamental to many
applications, such as intelligence analysis and decision assistance. However,
event extraction in the military field faces the data scarcity problem, which
impedes the research of event extraction models in this domain. To alleviate
this problem, we propose CMNEE, a large-scale, document-level open-source
Chinese Military News Event Extraction dataset. It contains 17,000 documents
and 29,223 events, which are all manually annotated based on a pre-defined
schema for the military domain including 8 event types and 11 argument role
types. We designed a two-stage, multi-turns annotation strategy to ensure the
quality of CMNEE and reproduced several state-of-the-art event extraction
models with a systematic evaluation. The experimental results on CMNEE fall
shorter than those on other domain datasets obviously, which demonstrates that
event extraction for military domain poses unique challenges and requires
further research efforts. Our code and data can be obtained from
https://github.com/Mzzzhu/CMNEE. | cs.CL | CMNEE: A Large-Scale Document-Level Event Extraction Dataset based on Open-Source Chinese Military News |
2024-04-18T15:01:00Z | http://arxiv.org/pdf/2404.12241v2 | http://arxiv.org/abs/2404.12241v2 | Bertie Vidgen, Adarsh Agrawal, Ahmed M. Ahmed, Victor Akinwande, Namir Al-Nuaimi, Najla Alfaraj, Elie Alhajjar, Lora Aroyo, Trupti Bavalatti, Max Bartolo, Borhane Blili-Hamelin, Kurt Bollacker, Rishi Bomassani, Marisa Ferrara Boston, Siméon Campos, Kal Chakra, Canyu Chen, Cody Coleman, Zacharie Delpierre Coudert, Leon Derczynski, Debojyoti Dutta, Ian Eisenberg, James Ezick, Heather Frase, Brian Fuller, Ram Gandikota, Agasthya Gangavarapu, Ananya Gangavarapu, James Gealy, Rajat Ghosh, James Goel, Usman Gohar, Sujata Goswami, Scott A. Hale, Wiebke Hutiri, Joseph Marvin Imperial, Surgan Jandial, Nick Judd, Felix Juefei-Xu, Foutse Khomh, Bhavya Kailkhura, Hannah Rose Kirk, Kevin Klyman, Chris Knotz, Michael Kuchnik, Shachi H. Kumar, Srijan Kumar, Chris Lengerich, Bo Li, Zeyi Liao, Eileen Peters Long, Victor Lu, Sarah Luger, Yifan Mai, Priyanka Mary Mammen, Kelvin Manyeki, Sean McGregor, Virendra Mehta, Shafee Mohammed, Emanuel Moss, Lama Nachman, Dinesh Jinenhally Naganna, Amin Nikanjam, Besmira Nushi, Luis Oala, Iftach Orr, Alicia Parrish, Cigdem Patlak, William Pietri, Forough Poursabzi-Sangdeh, Eleonora Presani, Fabrizio Puletti, Paul Röttger, Saurav Sahay, Tim Santos, Nino Scherrer, Alice Schoenauer Sebag, Patrick Schramowski, Abolfazl Shahbazi, Vin Sharma, Xudong Shen, Vamsi Sistla, Leonard Tang, Davide Testuggine, Vithursan Thangarasa, Elizabeth Anne Watkins, Rebecca Weiss, Chris Welty, Tyler Wilbers, Adina Williams, Carole-Jean Wu, Poonam Yadav, Xianjun Yang, Yi Zeng, Wenhui Zhang, Fedor Zhdanov, Jiacheng Zhu, Percy Liang, Peter Mattson, Joaquin Vanschoren | 2024-05-13T20:46:10Z | cs.CL, cs.AI | This paper introduces v0.5 of the AI Safety Benchmark, which has been created
by the MLCommons AI Safety Working Group. The AI Safety Benchmark has been
designed to assess the safety risks of AI systems that use chat-tuned language
models. We introduce a principled approach to specifying and constructing the
benchmark, which for v0.5 covers only a single use case (an adult chatting to a
general-purpose assistant in English), and a limited set of personas (i.e.,
typical users, malicious users, and vulnerable users). We created a new
taxonomy of 13 hazard categories, of which 7 have tests in the v0.5 benchmark.
We plan to release version 1.0 of the AI Safety Benchmark by the end of 2024.
The v1.0 benchmark will provide meaningful insights into the safety of AI
systems. However, the v0.5 benchmark should not be used to assess the safety of
AI systems. We have sought to fully document the limitations, flaws, and
challenges of v0.5. This release of v0.5 of the AI Safety Benchmark includes
(1) a principled approach to specifying and constructing the benchmark, which
comprises use cases, types of systems under test (SUTs), language and context,
personas, tests, and test items; (2) a taxonomy of 13 hazard categories with
definitions and subcategories; (3) tests for seven of the hazard categories,
each comprising a unique set of test items, i.e., prompts. There are 43,090
test items in total, which we created with templates; (4) a grading system for
AI systems against the benchmark; (5) an openly available platform, and
downloadable tool, called ModelBench that can be used to evaluate the safety of
AI systems on the benchmark; (6) an example evaluation report which benchmarks
the performance of over a dozen openly available chat-tuned language models;
(7) a test specification for the benchmark. | cs.CL | Introducing v0.5 of the AI Safety Benchmark from MLCommons |
2024-04-18T14:38:32Z | http://arxiv.org/pdf/2404.12224v2 | http://arxiv.org/abs/2404.12224v2 | Jie Wang, Tao Ji, Yuanbin Wu, Hang Yan, Tao Gui, Qi Zhang, Xuanjing Huang, Xiaoling Wang | 2024-05-28T01:38:59Z | cs.CL | Generalizing to longer sentences is important for recent Transformer-based
language models. Besides algorithms manipulating explicit position features,
the success of Transformers without position encodings (NoPE) provides a new
way to overcome the challenge. In this paper, we study the length
generalization property of NoPE. We find that although NoPE can extend to
longer sequences than the commonly used explicit position encodings, it still
has a limited context length. We identify a connection between the failure of
NoPE's generalization and the distraction of attention distributions. We
propose a parameter-efficient tuning for searching attention heads' best
temperature hyper-parameters, which substantially expands NoPE's context size.
Experiments on long sequence language modeling, the synthetic passkey retrieval
task and real-world long context tasks show that NoPE can achieve competitive
performances with state-of-the-art length generalization algorithms. The source
code is publicly accessible | cs.CL | Length Generalization of Causal Transformers without Position Encoding |
2024-04-18T13:57:18Z | http://arxiv.org/pdf/2404.12195v1 | http://arxiv.org/abs/2404.12195v1 | Chandeepa Dissanayake, Lahiru Lowe, Sachith Gunasekara, Yasiru Ratnayake | 2024-04-18T13:57:18Z | cs.CL, cs.LG | Instruction fine-tuning pretrained LLMs for diverse downstream tasks has
demonstrated remarkable success and has captured the interest of both academics
and practitioners. To ensure such fine-tuned LLMs align with human preferences,
techniques such as RLHF and DPO have emerged. At the same time, there is
increasing interest in smaller parameter counts for models. In this work, using
OpenLLaMA 3Bv2 as a base model, we describe the recipe used to fine-tune the
OpenBezoar family of models. In this recipe: We first generate synthetic
instruction fine-tuning data using an open and commercially non-restrictive
instruction fine-tuned variant of the Falcon-40B model under three schemes
based on: LaMini-LM, WizardLM/Evol-Instruct (with databricks-dolly-15k as a
seed dataset) and Orca (with the Flan Collection as a seed dataset), then
filter these generations using GPT-4 as a human proxy. We then perform
cost-effective QLoRA-based supervised fine-tuning sequentially with each
scheme. The resulting checkpoint is further fine-tuned with a subset of the
HH-RLHF dataset to minimize distribution shift prior to using the DPO loss to
obtain the final checkpoint. Evaluation is done with the LM Eval Harness
tasks/metrics as well as on MT-Bench using the "LLM-as-a-judge" framework with
Claude 2.1, with the finding that the final checkpoint,
"OpenBezoar-HH-RLHF-DPO", demonstrates superior performance over many models at
the 3B parameter scale, even outperforming the top model in one of the
categories on the Huggingface Open LLM Leaderboard. We release
"OpenBezoar-SFT", "OpenBezoar-HH-RLHF-SFT", "OpenBezoar-HH-RLHF-DPO"
checkpoints, alongside our generated datasets on HuggingFace at
https://huggingface.co/collections/SurgeGlobal/open-bezoar-6620a24923e12127e9e2b9cc
and our codebase at
https://bitbucket.org/paladinanalytics/workspace/projects/OP. | cs.CL | OpenBezoar: Small, Cost-Effective and Open Models Trained on Mixes of Instruction Data |
2024-04-18T13:31:57Z | http://arxiv.org/pdf/2404.12177v2 | http://arxiv.org/abs/2404.12177v2 | Aitor García-Pablos, Naiara Perez, Montse Cuadros, Jaione Bengoetxea | 2024-06-04T15:43:54Z | cs.CL | The widespread availability of Question Answering (QA) datasets in English
has greatly facilitated the advancement of the Natural Language Processing
(NLP) field. However, the scarcity of such resources for minority languages,
such as Basque, poses a substantial challenge for these communities. In this
context, the translation and alignment of existing QA datasets plays a crucial
role in narrowing this technological gap. This work presents EuSQuAD, the first
initiative dedicated to automatically translating and aligning SQuAD2.0 into
Basque, resulting in more than 142k QA examples. We demonstrate EuSQuAD's value
through extensive qualitative analysis and QA experiments supported with
EuSQuAD as training data. These experiments are evaluated with a new
human-annotated dataset. | cs.CL | EuSQuAD: Automatically Translated and Aligned SQuAD2.0 for Basque |
2024-04-18T13:31:05Z | http://arxiv.org/pdf/2404.12174v1 | http://arxiv.org/abs/2404.12174v1 | Laura Majer, Jan Šnajder | 2024-04-18T13:31:05Z | cs.CL | The increasing threat of disinformation calls for automating parts of the
fact-checking pipeline. Identifying text segments requiring fact-checking is
known as claim detection (CD) and claim check-worthiness detection (CW), the
latter incorporating complex domain-specific criteria of worthiness and often
framed as a ranking task. Zero- and few-shot LLM prompting is an attractive
option for both tasks, as it bypasses the need for labeled datasets and allows
verbalized claim and worthiness criteria to be directly used for prompting. We
evaluate the LLMs' predictive and calibration accuracy on five CD/CW datasets
from diverse domains, each utilizing a different worthiness criterion. We
investigate two key aspects: (1) how best to distill factuality and worthiness
criteria into a prompt and (2) what amount of context to provide for each
claim. To this end, we experiment with varying the level of prompt verbosity
and the amount of contextual information provided to the model. Our results
show that optimal prompt verbosity is domain-dependent, adding context does not
improve performance, and confidence scores can be directly used to produce
reliable check-worthiness rankings. | cs.CL | Claim Check-Worthiness Detection: How Well do LLMs Grasp Annotation Guidelines? |
2024-04-18T13:25:29Z | http://arxiv.org/pdf/2404.12171v1 | http://arxiv.org/abs/2404.12171v1 | İlker Gül, Rémi Lebret, Karl Aberer | 2024-04-18T13:25:29Z | cs.CL, cs.SI | Stance detection, a key task in natural language processing, determines an
author's viewpoint based on textual analysis. This study evaluates the
evolution of stance detection methods, transitioning from early machine
learning approaches to the groundbreaking BERT model, and eventually to modern
Large Language Models (LLMs) such as ChatGPT, LLaMa-2, and Mistral-7B. While
ChatGPT's closed-source nature and associated costs present challenges, the
open-source models like LLaMa-2 and Mistral-7B offers an encouraging
alternative. Initially, our research focused on fine-tuning ChatGPT, LLaMa-2,
and Mistral-7B using several publicly available datasets. Subsequently, to
provide a comprehensive comparison, we assess the performance of these models
in zero-shot and few-shot learning scenarios. The results underscore the
exceptional ability of LLMs in accurately detecting stance, with all tested
models surpassing existing benchmarks. Notably, LLaMa-2 and Mistral-7B
demonstrate remarkable efficiency and potential for stance detection, despite
their smaller sizes compared to ChatGPT. This study emphasizes the potential of
LLMs in stance detection and calls for more extensive research in this field. | cs.CL | Stance Detection on Social Media with Fine-Tuned Large Language Models |
2024-04-18T12:58:36Z | http://arxiv.org/pdf/2404.12152v1 | http://arxiv.org/abs/2404.12152v1 | Zunran Wang, Zhonghua Li, Wei Shen, Qi Ye, Liqiang Nie | 2024-04-18T12:58:36Z | cs.CL | Lexicon-based retrieval has gained siginificant popularity in text retrieval
due to its efficient and robust performance. To further enhance performance of
lexicon-based retrieval, researchers have been diligently incorporating
state-of-the-art methodologies like Neural retrieval and text-level contrastive
learning approaches. Nonetheless, despite the promising outcomes, current
lexicon-based retrieval methods have received limited attention in exploring
the potential benefits of feature context representations and term-level
knowledge guidance. In this paper, we introduce an innovative method by
introducing FEature Context and TErm-level Knowledge modules(FecTek). To
effectively enrich the feature context representations of term weight, the
Feature Context Module (FCM) is introduced, which leverages the power of BERT's
representation to determine dynamic weights for each element in the embedding.
Additionally, we develop a term-level knowledge guidance module (TKGM) for
effectively utilizing term-level knowledge to intelligently guide the modeling
process of term weight. Evaluation of the proposed method on MS Marco benchmark
demonstrates its superiority over the previous state-of-the-art approaches. | cs.CL | FecTek: Enhancing Term Weight in Lexicon-Based Retrieval with Feature Context and Term-level Knowledge |
2024-04-18T12:55:18Z | http://arxiv.org/pdf/2404.12150v1 | http://arxiv.org/abs/2404.12150v1 | Tomasz Korbak | 2024-04-18T12:55:18Z | cs.LG, cs.CL | Language models (LMs) trained on vast quantities of text data can acquire
sophisticated skills such as generating summaries, answering questions or
generating code. However, they also manifest behaviors that violate human
preferences, e.g., they can generate offensive content, falsehoods or
perpetuate social biases. In this thesis, I explore several approaches to
aligning LMs with human preferences. First, I argue that aligning LMs can be
seen as Bayesian inference: conditioning a prior (base, pretrained LM) on
evidence about human preferences (Chapter 2). Conditioning on human preferences
can be implemented in numerous ways. In Chapter 3, I investigate the relation
between two approaches to finetuning pretrained LMs using feedback given by a
scoring function: reinforcement learning from human feedback (RLHF) and
distribution matching. I show that RLHF can be seen as a special case of
distribution matching but distributional matching is strictly more general. In
chapter 4, I show how to extend the distribution matching to conditional
language models. Finally, in chapter 5 I explore a different root: conditioning
an LM on human preferences already during pretraining. I show that involving
human feedback from the very start tends to be more effective than using it
only during supervised finetuning. Overall, these results highlight the room
for alignment techniques different from and complementary to RLHF. | cs.LG | Aligning language models with human preferences |
2024-04-18T12:48:17Z | http://arxiv.org/pdf/2404.12145v1 | http://arxiv.org/abs/2404.12145v1 | Xenia Ohmer, Elia Bruni, Dieuwke Hupkes | 2024-04-18T12:48:17Z | cs.CL, cs.AI | The staggering pace with which the capabilities of large language models
(LLMs) are increasing, as measured by a range of commonly used natural language
understanding (NLU) benchmarks, raises many questions regarding what
"understanding" means for a language model and how it compares to human
understanding. This is especially true since many LLMs are exclusively trained
on text, casting doubt on whether their stellar benchmark performances are
reflective of a true understanding of the problems represented by these
benchmarks, or whether LLMs simply excel at uttering textual forms that
correlate with what someone who understands the problem would say. In this
philosophically inspired work, we aim to create some separation between form
and meaning, with a series of tests that leverage the idea that world
understanding should be consistent across presentational modes - inspired by
Fregean senses - of the same meaning. Specifically, we focus on consistency
across languages as well as paraphrases. Taking GPT-3.5 as our object of study,
we evaluate multisense consistency across five different languages and various
tasks. We start the evaluation in a controlled setting, asking the model for
simple facts, and then proceed with an evaluation on four popular NLU
benchmarks. We find that the model's multisense consistency is lacking and run
several follow-up analyses to verify that this lack of consistency is due to a
sense-dependent task understanding. We conclude that, in this aspect, the
understanding of LLMs is still quite far from being consistent and human-like,
and deliberate on how this impacts their utility in the context of learning
about human language and understanding. | cs.CL | From Form(s) to Meaning: Probing the Semantic Depths of Language Models Using Multisense Consistency |
2024-04-18T12:33:57Z | http://arxiv.org/pdf/2404.12132v1 | http://arxiv.org/abs/2404.12132v1 | Shahin Amiriparian, Maurice Gerczuk, Justina Lutz, Wolfgang Strube, Irina Papazova, Alkomiet Hasan, Alexander Kathan, Björn W. Schuller | 2024-04-18T12:33:57Z | cs.SD, cs.CL, eess.AS, I.2 | The delayed access to specialized psychiatric assessments and care for
patients at risk of suicidal tendencies in emergency departments creates a
notable gap in timely intervention, hindering the provision of adequate mental
health support during critical situations. To address this, we present a
non-invasive, speech-based approach for automatic suicide risk assessment. For
our study, we have collected a novel dataset of speech recordings from $20$
patients from which we extract three sets of features, including wav2vec,
interpretable speech and acoustic features, and deep learning-based spectral
representations. We proceed by conducting a binary classification to assess
suicide risk in a leave-one-subject-out fashion. Our most effective speech
model achieves a balanced accuracy of $66.2\,\%$. Moreover, we show that
integrating our speech model with a series of patients' metadata, such as the
history of suicide attempts or access to firearms, improves the overall result.
The metadata integration yields a balanced accuracy of $94.4\,\%$, marking an
absolute improvement of $28.2\,\%$, demonstrating the efficacy of our proposed
approaches for automatic suicide risk assessment in emergency medicine. | cs.SD | Enhancing Suicide Risk Assessment: A Speech-Based Automated Approach in Emergency Medicine |
2024-04-18T11:38:25Z | http://arxiv.org/pdf/2404.12104v1 | http://arxiv.org/abs/2404.12104v1 | Yuzhu Cai, Sheng Yin, Yuxi Wei, Chenxin Xu, Weibo Mao, Felix Juefei-Xu, Siheng Chen, Yanfeng Wang | 2024-04-18T11:38:25Z | cs.CV, cs.CL, cs.LG | The burgeoning landscape of text-to-image models, exemplified by innovations
such as Midjourney and DALLE 3, has revolutionized content creation across
diverse sectors. However, these advancements bring forth critical ethical
concerns, particularly with the misuse of open-source models to generate
content that violates societal norms. Addressing this, we introduce
Ethical-Lens, a framework designed to facilitate the value-aligned usage of
text-to-image tools without necessitating internal model revision. Ethical-Lens
ensures value alignment in text-to-image models across toxicity and bias
dimensions by refining user commands and rectifying model outputs. Systematic
evaluation metrics, combining GPT4-V, HEIM, and FairFace scores, assess
alignment capability. Our experiments reveal that Ethical-Lens enhances
alignment capabilities to levels comparable with or superior to commercial
models like DALLE 3, ensuring user-generated content adheres to ethical
standards while maintaining image quality. This study indicates the potential
of Ethical-Lens to ensure the sustainable development of open-source
text-to-image tools and their beneficial integration into society. Our code is
available at https://github.com/yuzhu-cai/Ethical-Lens. | cs.CV | Ethical-Lens: Curbing Malicious Usages of Open-Source Text-to-Image Models |
2024-04-18T11:29:23Z | http://arxiv.org/pdf/2404.12096v2 | http://arxiv.org/abs/2404.12096v2 | Dawei Zhu, Liang Wang, Nan Yang, Yifan Song, Wenhao Wu, Furu Wei, Sujian Li | 2024-04-25T02:26:15Z | cs.CL, cs.LG | Embedding models play a pivot role in modern NLP applications such as IR and
RAG. While the context limit of LLMs has been pushed beyond 1 million tokens,
embedding models are still confined to a narrow context window not exceeding 8k
tokens, refrained from application scenarios requiring long inputs such as
legal contracts. This paper explores context window extension of existing
embedding models, pushing the limit to 32k without requiring additional
training. First, we examine the performance of current embedding models for
long context retrieval on our newly constructed LongEmbed benchmark. LongEmbed
comprises two synthetic tasks and four carefully chosen real-world tasks,
featuring documents of varying length and dispersed target information.
Benchmarking results underscore huge room for improvement in these models.
Based on this, comprehensive experiments show that training-free context window
extension strategies like position interpolation can effectively extend the
context window of existing embedding models by several folds, regardless of
their original context being 512 or beyond 4k. Furthermore, for models
employing absolute position encoding (APE), we show the possibility of further
fine-tuning to harvest notable performance gains while strictly preserving
original behavior for short inputs. For models using rotary position embedding
(RoPE), significant enhancements are observed when employing RoPE-specific
methods, such as NTK and SelfExtend, indicating RoPE's superiority over APE for
context window extension. To facilitate future research, we release E5-Base-4k
and E5-RoPE-Base, along with the LongEmbed benchmark. | cs.CL | LongEmbed: Extending Embedding Models for Long Context Retrieval |
2024-04-18T10:59:54Z | http://arxiv.org/pdf/2404.12077v1 | http://arxiv.org/abs/2404.12077v1 | Rong Wang, Kun Sun | 2024-04-18T10:59:54Z | cs.SD, cs.AI, cs.CL, cs.LG, eess.AS | This study employs deep learning techniques to explore four speaker profiling
tasks on the TIMIT dataset, namely gender classification, accent
classification, age estimation, and speaker identification, highlighting the
potential and challenges of multi-task learning versus single-task models. The
motivation for this research is twofold: firstly, to empirically assess the
advantages and drawbacks of multi-task learning over single-task models in the
context of speaker profiling; secondly, to emphasize the undiminished
significance of skillful feature engineering for speaker recognition tasks. The
findings reveal challenges in accent classification, and multi-task learning is
found advantageous for tasks of similar complexity. Non-sequential features are
favored for speaker recognition, but sequential ones can serve as starting
points for complex models. The study underscores the necessity of meticulous
experimentation and parameter tuning for deep learning models. | cs.SD | TIMIT Speaker Profiling: A Comparison of Multi-task learning and Single-task learning Approaches |
2024-04-18T10:25:42Z | http://arxiv.org/pdf/2404.12065v1 | http://arxiv.org/abs/2404.12065v1 | M. Abdul Khaliq, P. Chang, M. Ma, B. Pflugfelder, F. Miletić | 2024-04-18T10:25:42Z | cs.CL, cs.AI, cs.CY, cs.ET, cs.MA | The escalating challenge of misinformation, particularly in the context of
political discourse, necessitates advanced solutions for fact-checking. We
introduce innovative approaches to enhance the reliability and efficiency of
multimodal fact-checking through the integration of Large Language Models
(LLMs) with Retrieval-augmented Generation (RAG)- based advanced reasoning
techniques. This work proposes two novel methodologies, Chain of RAG (CoRAG)
and Tree of RAG (ToRAG). The approaches are designed to handle multimodal
claims by reasoning the next questions that need to be answered based on
previous evidence. Our approaches improve the accuracy of veracity predictions
and the generation of explanations over the traditional fact-checking approach
of sub-question generation with chain of thought veracity prediction. By
employing multimodal LLMs adept at analyzing both text and images, this
research advances the capability of automated systems in identifying and
countering misinformation. | cs.CL | RAGAR, Your Falsehood RADAR: RAG-Augmented Reasoning for Political Fact-Checking using Multimodal Large Language Models |
2024-04-18T10:17:04Z | http://arxiv.org/pdf/2404.12059v1 | http://arxiv.org/abs/2404.12059v1 | Junjie Chen, Xiangheng He, Danushka Bollegala, Yusuke Miyao | 2024-04-18T10:17:04Z | cs.CL | Unsupervised constituency parsing is about identifying word sequences that
form a syntactic unit (i.e., constituents) in a target sentence. Linguists
identify the constituent by evaluating a set of Predicate-Argument Structure
(PAS) equivalent sentences where we find the constituent corresponds to
frequent word sequences. However, such information is unavailable to previous
parsing methods which identify the constituent by observing sentences with
diverse PAS. In this study, we empirically verify that \textbf{constituents
correspond to word sequence patterns in the PAS-equivalent sentence set}. We
propose a frequency-based method \emph{span-overlap}, applying the word
sequence pattern to computational unsupervised parsing for the first time.
Parsing experiments show that the span-overlap parser outperforms
state-of-the-art parsers in eight out of ten languages. Further discrimination
analysis confirms that the span-overlap method can non-trivially separate
constituents from non-constituents. This result highlights the utility of the
word sequence pattern. Additionally, we discover a multilingual phenomenon:
\textbf{participant-denoting constituents are more frequent than event-denoting
constituents}. The phenomenon indicates a behavioral difference between the two
constituent types, laying the foundation for future labeled unsupervised
parsing. | cs.CL | Constituents Correspond to Word Sequence Patterns among Sentences with Equivalent Predicate-Argument Structures: Unsupervised Constituency Parsing by Span Matching |
2024-04-18T10:06:00Z | http://arxiv.org/pdf/2404.12050v1 | http://arxiv.org/abs/2404.12050v1 | Jimenez Eladio, Hao Wu | 2024-04-18T10:06:00Z | cs.CL | Machine Reading Comprehension (MRC) holds a pivotal role in shaping Medical
Question Answering Systems (QAS) and transforming the landscape of accessing
and applying medical information. However, the inherent challenges in the
medical field, such as complex terminology and question ambiguity, necessitate
innovative solutions. One key solution involves integrating specialized medical
datasets and creating dedicated datasets. This strategic approach enhances the
accuracy of QAS, contributing to advancements in clinical decision-making and
medical research. To address the intricacies of medical terminology, a
specialized dataset was integrated, exemplified by a novel Span extraction
dataset derived from emrQA but restructured into 163,695 questions and 4,136
manually obtained answers, this new dataset was called emrQA-msquad dataset.
Additionally, for ambiguous questions, a dedicated medical dataset for the Span
extraction task was introduced, reinforcing the system's robustness. The
fine-tuning of models such as BERT, RoBERTa, and Tiny RoBERTa for medical
contexts significantly improved response accuracy within the F1-score range of
0.75 to 1.00 from 10.1% to 37.4%, 18.7% to 44.7% and 16.0% to 46.8%,
respectively. Finally, emrQA-msquad dataset is publicy available at
https://huggingface.co/datasets/Eladio/emrqa-msquad. | cs.CL | emrQA-msquad: A Medical Dataset Structured with the SQuAD V2.0 Framework, Enriched with emrQA Medical Information |
2024-04-18T09:58:51Z | http://arxiv.org/pdf/2404.12045v1 | http://arxiv.org/abs/2404.12045v1 | Jiaqi Li, Xiaobo Wang, Zihao Wang, Zilong Zheng | 2024-04-18T09:58:51Z | cs.AI, cs.CL | We introduce RAM, an innovative RAG-based framework with an ever-improving
memory. Inspired by humans' pedagogical process, RAM utilizes recursively
reasoning-based retrieval and experience reflections to continually update the
memory and learn from users' communicative feedback, namely communicative
learning. Extensive experiments with both simulated and real users demonstrate
significant improvements over traditional RAG and self-knowledge methods,
particularly excelling in handling false premise and multi-hop questions.
Furthermore, RAM exhibits promising adaptability to various feedback and
retrieval method chain types, showcasing its potential for advancing AI
capabilities in dynamic knowledge acquisition and lifelong learning. | cs.AI | RAM: Towards an Ever-Improving Memory System by Learning from Communications |
2024-04-18T09:52:50Z | http://arxiv.org/pdf/2404.12042v1 | http://arxiv.org/abs/2404.12042v1 | Abinew Ali Ayele, Esubalew Alemneh Jalew, Adem Chanie Ali, Seid Muhie Yimam, Chris Biemann | 2024-04-18T09:52:50Z | cs.CL | The prevalence of digital media and evolving sociopolitical dynamics have
significantly amplified the dissemination of hateful content. Existing studies
mainly focus on classifying texts into binary categories, often overlooking the
continuous spectrum of offensiveness and hatefulness inherent in the text. In
this research, we present an extensive benchmark dataset for Amharic,
comprising 8,258 tweets annotated for three distinct tasks: category
classification, identification of hate targets, and rating offensiveness and
hatefulness intensities. Our study highlights that a considerable majority of
tweets belong to the less offensive and less hate intensity levels,
underscoring the need for early interventions by stakeholders. The prevalence
of ethnic and political hatred targets, with significant overlaps in our
dataset, emphasizes the complex relationships within Ethiopia's sociopolitical
landscape. We build classification and regression models and investigate the
efficacy of models in handling these tasks. Our results reveal that hate and
offensive speech can not be addressed by a simplistic binary classification,
instead manifesting as variables across a continuous range of values. The
Afro-XLMR-large model exhibits the best performances achieving F1-scores of
75.30%, 70.59%, and 29.42% for the category, target, and regression tasks,
respectively. The 80.22% correlation coefficient of the Afro-XLMR-large model
indicates strong alignments. | cs.CL | Exploring Boundaries and Intensities in Offensive and Hate Speech: Unveiling the Complex Spectrum of Social Media Discourse |
2024-04-18T09:52:18Z | http://arxiv.org/pdf/2404.12041v2 | http://arxiv.org/abs/2404.12041v2 | Siya Qi, Yulan He, Zheng Yuan | 2024-06-15T22:57:20Z | cs.CL, cs.AI | Hallucination in Natural Language Generation (NLG) is like the elephant in
the room, obvious but often overlooked until recent achievements significantly
improved the fluency and grammaticality of generated text. As the capabilities
of text generation models have improved, researchers have begun to pay more
attention to the phenomenon of hallucination. Despite significant progress in
this field in recent years, the evaluation system for hallucination is complex
and diverse, lacking clear organization. We are the first to comprehensively
survey how various evaluation methods have evolved with the development of text
generation models from three dimensions, including hallucinated fact
granularity, evaluator design principles, and assessment facets. This survey
aims to help researchers identify current limitations in hallucination
evaluation and highlight future research directions. | cs.CL | Can We Catch the Elephant? A Survey of the Evolvement of Hallucination Evaluation on Natural Language Generation |
2024-04-18T09:46:25Z | http://arxiv.org/pdf/2404.12038v3 | http://arxiv.org/abs/2404.12038v3 | Zhihao Xu, Ruixuan Huang, Changyu Chen, Shuai Wang, Xiting Wang | 2024-07-02T14:17:57Z | cs.CL | Despite careful safety alignment, current large language models (LLMs) remain
vulnerable to various attacks. To further unveil the safety risks of LLMs, we
introduce a Safety Concept Activation Vector (SCAV) framework, which
effectively guides the attacks by accurately interpreting LLMs' safety
mechanisms. We then develop an SCAV-guided attack method that can generate both
attack prompts and embedding-level attacks with automatically selected
perturbation hyperparameters. Both automatic and human evaluations demonstrate
that our attack method significantly improves the attack success rate and
response quality while requiring less training data. Additionally, we find that
our generated attack prompts may be transferable to GPT-4, and the
embedding-level attacks may also be transferred to other white-box LLMs whose
parameters are known. Our experiments further uncover the safety risks present
in current LLMs. For example, we find that six out of seven open-source LLMs
that we attack consistently provide relevant answers to more than 85\%
malicious instructions. Finally, we provide insights into the safety mechanism
of LLMs. | cs.CL | Uncovering Safety Risks of Large Language Models through Concept Activation Vector |
2024-04-18T09:17:06Z | http://arxiv.org/pdf/2404.12022v1 | http://arxiv.org/abs/2404.12022v1 | Pengfei Wu, Jiahao Liu, Zhuocheng Gong, Qifan Wang, Jinpeng Li, Jingang Wang, Xunliang Cai, Dongyan Zhao | 2024-04-18T09:17:06Z | cs.CL | Large language models (LLMs) have recently shown remarkable performance
across a wide range of tasks. However, the substantial number of parameters in
LLMs contributes to significant latency during model inference. This is
particularly evident when utilizing autoregressive decoding methods, which
generate one token in a single forward process, thereby not fully capitalizing
on the parallel computing capabilities of GPUs. In this paper, we propose a
novel parallel decoding approach, namely \textit{hidden transfer}, which
decodes multiple successive tokens simultaneously in a single forward pass. The
idea is to transfer the intermediate hidden states of the previous context to
the \textit{pseudo} hidden states of the future tokens to be generated, and
then the pseudo hidden states will pass the following transformer layers
thereby assimilating more semantic information and achieving superior
predictive accuracy of the future tokens.
Besides, we use the novel tree attention mechanism to simultaneously generate
and verify multiple candidates of output sequences, which ensure the lossless
generation and further improves the generation efficiency of our method.
Experiments demonstrate the effectiveness of our method. We conduct a lot of
analytic experiments to prove our motivation. In terms of acceleration metrics,
we outperform all the single-model acceleration techniques, including Medusa
and Self-Speculative decoding. | cs.CL | Parallel Decoding via Hidden Transfer for Lossless Large Language Model Acceleration |
2024-04-18T09:04:39Z | http://arxiv.org/pdf/2404.12014v1 | http://arxiv.org/abs/2404.12014v1 | Zi Xiong, Lizhi Qing, Yangyang Kang, Jiawei Liu, Hongsong Li, Changlong Sun, Xiaozhong Liu, Wei Lu | 2024-04-18T09:04:39Z | cs.CL, cs.CR | The widespread use of pre-trained language models (PLMs) in natural language
processing (NLP) has greatly improved performance outcomes. However, these
models' vulnerability to adversarial attacks (e.g., camouflaged hints from drug
dealers), particularly in the Chinese language with its rich character
diversity/variation and complex structures, hatches vital apprehension. In this
study, we propose a novel method, CHinese vAriatioN Graph Enhancement (CHANGE),
to increase the robustness of PLMs against character variation attacks in
Chinese content. CHANGE presents a novel approach for incorporating a Chinese
character variation graph into the PLMs. Through designing different
supplementary tasks utilizing the graph structure, CHANGE essentially enhances
PLMs' interpretation of adversarially manipulated text. Experiments conducted
in a multitude of NLP tasks show that CHANGE outperforms current language
models in combating against adversarial attacks and serves as a valuable
contribution to robust language model research. These findings contribute to
the groundwork on robust language models and highlight the substantial
potential of graph-guided pre-training strategies for real-world applications. | cs.CL | Enhance Robustness of Language Models Against Variation Attack through Graph Integration |
2024-04-18T09:04:15Z | http://arxiv.org/pdf/2404.12013v1 | http://arxiv.org/abs/2404.12013v1 | Semih Yagcioglu, Osman Batur İnce, Aykut Erdem, Erkut Erdem, Desmond Elliott, Deniz Yuret | 2024-04-18T09:04:15Z | cs.CL | The rise of large-scale multimodal models has paved the pathway for
groundbreaking advances in generative modeling and reasoning, unlocking
transformative applications in a variety of complex tasks. However, a pressing
question that remains is their genuine capability for stronger forms of
generalization, which has been largely underexplored in the multimodal setting.
Our study aims to address this by examining sequential compositional
generalization using \textsc{CompAct} (\underline{Comp}ositional
\underline{Act}ivities)\footnote{Project Page:
\url{http://cyberiada.github.io/CompAct}}, a carefully constructed,
perceptually grounded dataset set within a rich backdrop of egocentric kitchen
activity videos. Each instance in our dataset is represented with a combination
of raw video footage, naturally occurring sound, and crowd-sourced step-by-step
descriptions. More importantly, our setup ensures that the individual concepts
are consistently distributed across training and evaluation sets, while their
compositions are novel in the evaluation set. We conduct a comprehensive
assessment of several unimodal and multimodal models. Our findings reveal that
bi-modal and tri-modal models exhibit a clear edge over their text-only
counterparts. This highlights the importance of multimodality while charting a
trajectory for future research in this domain. | cs.CL | Sequential Compositional Generalization in Multimodal Models |
2024-04-18T09:02:45Z | http://arxiv.org/pdf/2404.12010v1 | http://arxiv.org/abs/2404.12010v1 | Lasal Jayawardena, Prasan Yapa | 2024-04-18T09:02:45Z | cs.CL, cs.AI, cs.LG | Paraphrase generation is a pivotal task in natural language processing (NLP).
Existing datasets in the domain lack syntactic and lexical diversity, resulting
in paraphrases that closely resemble the source sentences. Moreover, these
datasets often contain hate speech and noise, and may unintentionally include
non-English language sentences. This research introduces ParaFusion, a
large-scale, high-quality English paraphrase dataset developed using Large
Language Models (LLM) to address these challenges. ParaFusion augments existing
datasets with high-quality data, significantly enhancing both lexical and
syntactic diversity while maintaining close semantic similarity. It also
mitigates the presence of hate speech and reduces noise, ensuring a cleaner and
more focused English dataset. Results show that ParaFusion offers at least a
25% improvement in both syntactic and lexical diversity, measured across
several metrics for each data source. The paper also aims to set a gold
standard for paraphrase evaluation as it contains one of the most comprehensive
evaluation strategies to date. The results underscore the potential of
ParaFusion as a valuable resource for improving NLP applications. | cs.CL | ParaFusion: A Large-Scale LLM-Driven English Paraphrase Dataset Infused with High-Quality Lexical and Syntactic Diversity |
2024-04-18T08:56:47Z | http://arxiv.org/pdf/2404.12006v1 | http://arxiv.org/abs/2404.12006v1 | Qian Li, Cheng Ji, Shu Guo, Yong Zhao, Qianren Mao, Shangguang Wang, Yuntao Wei, Jianxin Li | 2024-04-18T08:56:47Z | cs.CL | Multi-modal relation extraction (MMRE) is a challenging task that aims to
identify relations between entities in text leveraging image information.
Existing methods are limited by their neglect of the multiple entity pairs in
one sentence sharing very similar contextual information (ie, the same text and
image), resulting in increased difficulty in the MMRE task. To address this
limitation, we propose the Variational Multi-Modal Hypergraph Attention Network
(VM-HAN) for multi-modal relation extraction. Specifically, we first construct
a multi-modal hypergraph for each sentence with the corresponding image, to
establish different high-order intra-/inter-modal correlations for different
entity pairs in each sentence. We further design the Variational Hypergraph
Attention Networks (V-HAN) to obtain representational diversity among different
entity pairs using Gaussian distribution and learn a better hypergraph
structure via variational attention. VM-HAN achieves state-of-the-art
performance on the multi-modal relation extraction task, outperforming existing
methods in terms of accuracy and efficiency. | cs.CL | Variational Multi-Modal Hypergraph Attention Network for Multi-Modal Relation Extraction |
2024-04-18T08:49:38Z | http://arxiv.org/pdf/2404.11999v4 | http://arxiv.org/abs/2404.11999v4 | Yongcheng Zeng, Guoqing Liu, Weiyu Ma, Ning Yang, Haifeng Zhang, Jun Wang | 2024-06-27T15:27:41Z | cs.CL, cs.AI | Fine-tuning pre-trained Large Language Models (LLMs) is essential to align
them with human values and intentions. This process often utilizes methods like
pairwise comparisons and KL divergence against a reference LLM, focusing on the
evaluation of full answers generated by the models. However, the generation of
these responses occurs in a token level, following a sequential,
auto-regressive fashion. In this paper, we introduce Token-level Direct
Preference Optimization (TDPO), a novel approach to align LLMs with human
preferences by optimizing policy at the token level. Unlike previous methods,
which face challenges in divergence efficiency, TDPO incorporates forward KL
divergence constraints for each token, improving alignment and diversity.
Utilizing the Bradley-Terry model for a token-based reward system, TDPO
enhances the regulation of KL divergence, while preserving simplicity without
the need for explicit reward modeling. Experimental results across various text
tasks demonstrate TDPO's superior performance in balancing alignment with
generation diversity. Notably, fine-tuning with TDPO strikes a better balance
than DPO in the controlled sentiment generation and single-turn dialogue
datasets, and significantly improves the quality of generated responses
compared to both DPO and PPO-based RLHF methods. Our code is open-sourced at
https://github.com/Vance0124/Token-level-Direct-Preference-Optimization. | cs.CL | Token-level Direct Preference Optimization |
2024-04-18T08:14:53Z | http://arxiv.org/pdf/2404.11978v1 | http://arxiv.org/abs/2404.11978v1 | Zhengwei Tao, Xiancai Chen, Zhi Jin, Xiaoying Bai, Haiyan Zhao, Yiwei Lou | 2024-04-18T08:14:53Z | cs.CL | Events refer to specific occurrences, incidents, or happenings that take
place under a particular background. Event reasoning aims to infer events
according to certain relations and predict future events. The cutting-edge
techniques for event reasoning play a crucial role in various natural language
processing applications. Large language models (LLMs) have made significant
advancements in event reasoning owing to their wealth of knowledge and
reasoning capabilities. However, smaller instruction-tuned models currently in
use do not consistently demonstrate exceptional proficiency in managing these
tasks. This discrepancy arises from the absence of explicit modeling of events
and the interconnections of them within their instruction data. Consequently,
these models face challenges in comprehending event structures and semantics
while struggling to bridge the gap between their interpretations and human
understanding of events. Additionally, their limitations in grasping event
relations lead to constrained event reasoning abilities to effectively deduce
and incorporate pertinent event knowledge. In this paper, we propose
Event-Oriented Instruction Tuning (EvIT) to train our LLM. Specifically, we
first propose a novel structure named event quadruple which contains the
structure and semantics of events and is complete in the event representation.
We then design event-relation learning based on the structures. We encapsulate
the learning into the instruction-tuning formulation to better stimulate the
event reasoning capacity of our model. We design a heuristic unsupervised
method to mine event quadruple from a large-scale corpus. At last, we finetune
a Llama model on our Event-Oriented Instruction Tuning. We conduct extensive
experiments on event reasoning tasks on several datasets. Automatic and human
evaluations demonstrate EvIT achieves competitive performances on event
reasoning. | cs.CL | EVIT: Event-Oriented Instruction Tuning for Event Reasoning |
2024-04-18T07:59:53Z | http://arxiv.org/pdf/2404.11972v2 | http://arxiv.org/abs/2404.11972v2 | Hyuhng Joon Kim, Youna Kim, Cheonbok Park, Junyeob Kim, Choonghyun Park, Kang Min Yoo, Sang-goo Lee, Taeuk Kim | 2024-06-17T03:04:32Z | cs.CL | In interactions between users and language model agents, user utterances
frequently exhibit ellipsis (omission of words or phrases) or imprecision (lack
of exactness) to prioritize efficiency. This can lead to varying
interpretations of the same input based on different assumptions or background
knowledge. It is thus crucial for agents to adeptly handle the inherent
ambiguity in queries to ensure reliability. However, even state-of-the-art
large language models (LLMs) still face challenges in such scenarios, primarily
due to the following hurdles: (1) LLMs are not explicitly trained to deal with
ambiguous utterances; (2) the degree of ambiguity perceived by the LLMs may
vary depending on the possessed knowledge. To address these issues, we propose
Alignment with Perceived Ambiguity (APA), a novel pipeline that aligns LLMs to
manage ambiguous queries by leveraging their own assessment of ambiguity (i.e.,
perceived ambiguity). Experimental results on question-answering datasets
demonstrate that APA empowers LLMs to explicitly detect and manage ambiguous
queries while retaining the ability to answer clear questions. Furthermore, our
finding proves that APA excels beyond training with gold-standard labels,
especially in out-of-distribution scenarios. | cs.CL | Aligning Language Models to Explicitly Handle Ambiguity |
2024-04-18T07:55:02Z | http://arxiv.org/pdf/2404.11968v1 | http://arxiv.org/abs/2404.11968v1 | Chuanhao Xu, Jingwei Cheng, Fu Zhang | 2024-04-18T07:55:02Z | cs.CL, I.2.4 | Entity alignment (EA) aims to find equivalent entities between two Knowledge
Graphs. Existing embedding-based EA methods usually encode entities as
embeddings, triples as embeddings' constraint and learn to align the
embeddings. The structural and side information are usually utilized via
embedding propagation, aggregation or interaction. However, the details of the
underlying logical inference steps among the alignment process are usually
omitted, resulting in inadequate inference process. In this paper, we introduce
P-NAL, an entity alignment method that captures two types of logical inference
paths with Non-Axiomatic Logic (NAL). Type 1 is the bridge-like inference path
between to-be-aligned entity pairs, consisting of two relation/attribute
triples and a similarity sentence between the other two entities. Type 2 links
the entity pair by their embeddings. P-NAL iteratively aligns entities and
relations by integrating the conclusions of the inference paths. Moreover, our
method is logically interpretable and extensible due to the expressiveness of
NAL. Our proposed method is suitable for various EA settings. Experimental
results show that our method outperforms state-of-the-art methods in terms of
Hits@1, achieving 0.98+ on all three datasets of DBP15K with both supervised
and unsupervised settings. To our knowledge, we present the first in-depth
analysis of entity alignment's basic principles from a unified logical
perspective. | cs.CL | P-NAL: an Effective and Interpretable Entity Alignment Method |
2024-04-18T06:20:50Z | http://arxiv.org/pdf/2404.11932v2 | http://arxiv.org/abs/2404.11932v2 | Geyu Lin, Bin Wang, Zhengyuan Liu, Nancy F. Chen | 2024-06-12T09:35:48Z | cs.CL, cs.AI | Multilingual proficiency presents a significant challenge for large language
models (LLMs). English-centric models are usually suboptimal in other
languages, particularly those that are linguistically distant from English.
This performance discrepancy mainly stems from the imbalanced distribution of
training data across languages during pre-training and instruction tuning
stages. To address this problem, we propose a novel approach called CrossIn,
which utilizes a mixed composition of cross-lingual instruction tuning data.
Our method leverages the compressed representation shared by various languages
to efficiently enhance the model's task-solving capabilities and multilingual
proficiency within a single process. In addition, we introduce a multi-task and
multi-faceted benchmark to evaluate the effectiveness of CrossIn. Experimental
results demonstrate that our method substantially improves performance across
tasks and languages, and we provide extensive insights into the impact of
cross-lingual data volume and the integration of translation data on enhancing
multilingual consistency and accuracy. | cs.CL | CrossIn: An Efficient Instruction Tuning Approach for Cross-Lingual Knowledge Alignment |
2024-04-18T05:43:50Z | http://arxiv.org/pdf/2404.11916v1 | http://arxiv.org/abs/2404.11916v1 | Nakyeong Yang, Junseok Kim, Jiwon Moon, Yunah Jang, Kyomin Jung | 2024-04-18T05:43:50Z | cs.CL, cs.AI | Prompt-tuning methods have shown comparable performance as
parameter-efficient fine-tuning (PEFT) methods in various natural language
understanding tasks. However, existing prompt tuning methods still utilize the
entire model architecture; thus, they fail to accelerate inference speed in the
application. In this paper, we propose a novel approach called SKIll-localized
Prompt tuning (SKIP), which is extremely efficient in inference time. Our
method significantly enhances inference efficiency by investigating and
utilizing a skill-localized subnetwork in a language model. Surprisingly, our
method improves the inference speed up to 160% while pruning 52% of the
parameters. Furthermore, we demonstrate that our method is applicable across
various transformer-based architectures, thereby confirming its practicality
and scalability. | cs.CL | SKIP: Skill-Localized Prompt Tuning for Inference Speed Boost-Up |
2024-04-18T05:25:54Z | http://arxiv.org/pdf/2404.11912v2 | http://arxiv.org/abs/2404.11912v2 | Hanshi Sun, Zhuoming Chen, Xinyu Yang, Yuandong Tian, Beidi Chen | 2024-04-23T03:38:13Z | cs.CL, cs.LG | With large language models (LLMs) widely deployed in long content generation
recently, there has emerged an increasing demand for efficient long-sequence
inference support. However, key-value (KV) cache, which is stored to avoid
re-computation, has emerged as a critical bottleneck by growing linearly in
size with the sequence length. Due to the auto-regressive nature of LLMs, the
entire KV cache will be loaded for every generated token, resulting in low
utilization of computational cores and high latency. While various compression
methods for KV cache have been proposed to alleviate this issue, they suffer
from degradation in generation quality. We introduce TriForce, a hierarchical
speculative decoding system that is scalable to long sequence generation. This
approach leverages the original model weights and dynamic sparse KV cache via
retrieval as a draft model, which serves as an intermediate layer in the
hierarchy and is further speculated by a smaller model to reduce its drafting
latency. TriForce not only facilitates impressive speedups for Llama2-7B-128K,
achieving up to 2.31$\times$ on an A100 GPU but also showcases scalability in
handling even longer contexts. For the offloading setting on two RTX 4090 GPUs,
TriForce achieves 0.108s/token$\unicode{x2014}$only half as slow as the
auto-regressive baseline on an A100, which attains 7.78$\times$ on our
optimized offloading system. Additionally, TriForce performs 4.86$\times$ than
DeepSpeed-Zero-Inference on a single RTX 4090 GPU. TriForce's robustness is
highlighted by its consistently outstanding performance across various
temperatures. The code is available at
https://github.com/Infini-AI-Lab/TriForce. | cs.CL | TriForce: Lossless Acceleration of Long Sequence Generation with Hierarchical Speculative Decoding |
2024-04-18T04:36:37Z | http://arxiv.org/pdf/2404.11891v1 | http://arxiv.org/abs/2404.11891v1 | Yilun Hao, Yongchao Chen, Yang Zhang, Chuchu Fan | 2024-04-18T04:36:37Z | cs.AI, cs.CL, cs.HC | The recent advancements of Large Language Models (LLMs), with their abundant
world knowledge and capabilities of tool-using and reasoning, fostered many LLM
planning algorithms. However, LLMs have not shown to be able to accurately
solve complex combinatorial optimization problems. In Xie et al. (2024), the
authors proposed TravelPlanner, a U.S. domestic travel planning benchmark, and
showed that LLMs themselves cannot make travel plans that satisfy user
requirements with a best success rate of 0.6%. In this work, we propose a
framework that enables LLMs to formally formulate and solve the travel planning
problem as a satisfiability modulo theory (SMT) problem and use SMT solvers
interactively and automatically solve the combinatorial search problem. The SMT
solvers guarantee the satisfiable of input constraints and the LLMs can enable
a language-based interaction with our framework. When the input constraints
cannot be satisfiable, our LLM-based framework will interactively offer
suggestions to users to modify their travel requirements via automatic
reasoning using the SMT solvers. We evaluate our framework with TravelPlanner
and achieve a success rate of 97%. We also create a separate dataset that
contain international travel benchmarks and use both dataset to evaluate the
effectiveness of our interactive planning framework when the initial user
queries cannot be satisfied. Our framework could generate valid plans with an
average success rate of 78.6% for our dataset and 85.0% for TravelPlanner
according to diverse humans preferences. | cs.AI | Large Language Models Can Plan Your Travels Rigorously with Formal Verification Tools |
2024-04-18T03:03:46Z | http://arxiv.org/pdf/2404.11870v1 | http://arxiv.org/abs/2404.11870v1 | Hung Le, Dung Nguyen, Kien Do, Svetha Venkatesh, Truyen Tran | 2024-04-18T03:03:46Z | cs.LG, cs.CL | We propose Pointer-Augmented Neural Memory (PANM) to help neural networks
understand and apply symbol processing to new, longer sequences of data. PANM
integrates an external neural memory that uses novel physical addresses and
pointer manipulation techniques to mimic human and computer symbol processing
abilities. PANM facilitates pointer assignment, dereference, and arithmetic by
explicitly using physical pointers to access memory content. Remarkably, it can
learn to perform these operations through end-to-end training on sequence data,
powering various sequential models. Our experiments demonstrate PANM's
exceptional length extrapolating capabilities and improved performance in tasks
that require symbol processing, such as algorithmic reasoning and Dyck language
recognition. PANM helps Transformer achieve up to 100% generalization accuracy
in compositional learning tasks and significantly better results in
mathematical reasoning, question answering and machine translation tasks. | cs.LG | Enhancing Length Extrapolation in Sequential Models with Pointer-Augmented Neural Memory |
2024-04-18T01:48:28Z | http://arxiv.org/pdf/2404.11845v1 | http://arxiv.org/abs/2404.11845v1 | Isar Nejadgholi, Kathleen C. Fraser, Anna Kerkhof, Svetlana Kiritchenko | 2024-04-18T01:48:28Z | cs.CL, cs.CY | Gender stereotypes are pervasive beliefs about individuals based on their
gender that play a significant role in shaping societal attitudes, behaviours,
and even opportunities. Recognizing the negative implications of gender
stereotypes, particularly in online communications, this study investigates
eleven strategies to automatically counter-act and challenge these views. We
present AI-generated gender-based counter-stereotypes to (self-identified) male
and female study participants and ask them to assess their offensiveness,
plausibility, and potential effectiveness. The strategies of counter-facts and
broadening universals (i.e., stating that anyone can have a trait regardless of
group membership) emerged as the most robust approaches, while humour,
perspective-taking, counter-examples, and empathy for the speaker were
perceived as less effective. Also, the differences in ratings were more
pronounced for stereotypes about the different targets than between the genders
of the raters. Alarmingly, many AI-generated counter-stereotypes were perceived
as offensive and/or implausible. Our analysis and the collected dataset offer
foundational insight into counter-stereotype generation, guiding future efforts
to develop strategies that effectively challenge gender stereotypes in online
interactions. | cs.CL | Challenging Negative Gender Stereotypes: A Study on the Effectiveness of Automated Counter-Stereotypes |
2024-04-18T01:15:41Z | http://arxiv.org/pdf/2404.11826v1 | http://arxiv.org/abs/2404.11826v1 | Minbeom Kim, Hwanhee Lee, Joonsuk Park, Hwaran Lee, Kyomin Jung | 2024-04-18T01:15:41Z | cs.CL | As the integration of large language models into daily life is on the rise,
there is a clear gap in benchmarks for advising on subjective and personal
dilemmas. To address this, we introduce AdvisorQA, the first benchmark
developed to assess LLMs' capability in offering advice for deeply personalized
concerns, utilizing the LifeProTips subreddit forum. This forum features a
dynamic interaction where users post advice-seeking questions, receiving an
average of 8.9 advice per query, with 164.2 upvotes from hundreds of users,
embodying a collective intelligence framework. Therefore, we've completed a
benchmark encompassing daily life questions, diverse corresponding responses,
and majority vote ranking to train our helpfulness metric. Baseline experiments
validate the efficacy of AdvisorQA through our helpfulness metric, GPT-4, and
human evaluation, analyzing phenomena beyond the trade-off between helpfulness
and harmlessness. AdvisorQA marks a significant leap in enhancing QA systems
for providing personalized, empathetic advice, showcasing LLMs' improved
understanding of human subjectivity. | cs.CL | AdvisorQA: Towards Helpful and Harmless Advice-seeking Question Answering with Collective Intelligence |
2024-04-18T00:20:48Z | http://arxiv.org/pdf/2405.02318v1 | http://arxiv.org/abs/2405.02318v1 | Abhinav Lalwani, Lovish Chopra, Christopher Hahn, Caroline Trippel, Zhijing Jin, Mrinmaya Sachan | 2024-04-18T00:20:48Z | cs.CL, cs.AI, cs.LG, cs.LO | Logical fallacies are common errors in reasoning that undermine the logic of
an argument. Automatically detecting logical fallacies has important
applications in tracking misinformation and validating claims. In this paper,
we design a process to reliably detect logical fallacies by translating natural
language to First-order Logic (FOL) step-by-step using Large Language Models
(LLMs). We then utilize Satisfiability Modulo Theory (SMT) solvers to reason
about the validity of the formula and classify inputs as either a fallacy or
valid statement. Our model also provides a novel means of utilizing LLMs to
interpret the output of the SMT solver, offering insights into the
counter-examples that illustrate why a given sentence is considered a logical
fallacy. Our approach is robust, interpretable and does not require training
data or fine-tuning. We evaluate our model on a mixed dataset of fallacies and
valid sentences. The results demonstrate improved performance compared to
end-to-end LLMs, with our classifier achieving an F1-score of 71\% on the Logic
dataset. The approach is able to generalize effectively, achieving an F1-score
of 73% on the challenge set, LogicClimate, outperforming state-of-the-art
models by 21% despite its much smaller size. | cs.CL | NL2FOL: Translating Natural Language to First-Order Logic for Logical Fallacy Detection |
2024-04-18T00:05:02Z | http://arxiv.org/pdf/2404.11809v1 | http://arxiv.org/abs/2404.11809v1 | Xincan Feng, Zhi Qu, Yuchang Cheng, Taro Watanabe, Nobuhiro Yugami | 2024-04-18T00:05:02Z | cs.CL, cs.LG | A Knowledge Graph (KG) is the directed graphical representation of entities
and relations in the real world. KG can be applied in diverse Natural Language
Processing (NLP) tasks where knowledge is required. The need to scale up and
complete KG automatically yields Knowledge Graph Embedding (KGE), a shallow
machine learning model that is suffering from memory and training time
consumption issues. To mitigate the computational load, we propose a
parameter-sharing method, i.e., using conjugate parameters for complex numbers
employed in KGE models. Our method improves memory efficiency by 2x in relation
embedding while achieving comparable performance to the state-of-the-art
non-conjugate models, with faster, or at least comparable, training time. We
demonstrated the generalizability of our method on two best-performing KGE
models $5^{\bigstar}\mathrm{E}$ and $\mathrm{ComplEx}$ on five benchmark
datasets. | cs.CL | Sharing Parameter by Conjugation for Knowledge Graph Embeddings in Complex Space |
2024-04-17T23:00:29Z | http://arxiv.org/pdf/2404.11793v1 | http://arxiv.org/abs/2404.11793v1 | Mohammad Khosravani, Chenyang Huang, Amine Trabelsi | 2024-04-17T23:00:29Z | cs.CL | The proliferation of social media platforms has given rise to the amount of
online debates and arguments. Consequently, the need for automatic
summarization methods for such debates is imperative, however this area of
summarization is rather understudied. The Key Point Analysis (KPA) task
formulates argument summarization as representing the summary of a large
collection of arguments in the form of concise sentences in bullet-style
format, called key points. A sub-task of KPA, called Key Point Generation
(KPG), focuses on generating these key points given the arguments. This paper
introduces a novel extractive approach for key point generation, that
outperforms previous state-of-the-art methods for the task. Our method utilizes
an extractive clustering based approach that offers concise, high quality
generated key points with higher coverage of reference summaries, and less
redundant outputs. In addition, we show that the existing evaluation metrics
for summarization such as ROUGE are incapable of differentiating between
generated key points of different qualities. To this end, we propose a new
evaluation metric for assessing the generated key points by their coverage. Our
code can be accessed online. | cs.CL | Enhancing Argument Summarization: Prioritizing Exhaustiveness in Key Point Generation and Introducing an Automatic Coverage Evaluation Metric |
2024-04-17T22:12:41Z | http://arxiv.org/pdf/2404.11782v1 | http://arxiv.org/abs/2404.11782v1 | Sana Ebrahimi, Nima Shahbazi, Abolfazl Asudeh | 2024-04-17T22:12:41Z | cs.CL, cs.AI, cs.CY, cs.LG | The extensive scope of large language models (LLMs) across various domains
underscores the critical importance of responsibility in their application,
beyond natural language processing. In particular, the randomized nature of
LLMs, coupled with inherent biases and historical stereotypes in data, raises
critical concerns regarding reliability and equity. Addressing these challenges
are necessary before using LLMs for applications with societal impact. Towards
addressing this gap, we introduce REQUAL-LM, a novel method for finding
reliable and equitable LLM outputs through aggregation. Specifically, we
develop a Monte Carlo method based on repeated sampling to find a reliable
output close to the mean of the underlying distribution of possible outputs. We
formally define the terms such as reliability and bias, and design an
equity-aware aggregation to minimize harmful bias while finding a highly
reliable output. REQUAL-LM does not require specialized hardware, does not
impose a significant computing load, and uses LLMs as a blackbox. This design
choice enables seamless scalability alongside the rapid advancement of LLM
technologies. Our system does not require retraining the LLMs, which makes it
deployment ready and easy to adapt. Our comprehensive experiments using various
tasks and datasets demonstrate that REQUAL- LM effectively mitigates bias and
selects a more equitable response, specifically the outputs that properly
represents minority groups. | cs.CL | REQUAL-LM: Reliability and Equity through Aggregation in Large Language Models |
2024-04-17T21:27:33Z | http://arxiv.org/pdf/2404.11757v1 | http://arxiv.org/abs/2404.11757v1 | Mike A. Merrill, Mingtian Tan, Vinayak Gupta, Tom Hartvigsen, Tim Althoff | 2024-04-17T21:27:33Z | cs.CL | Time series are critical for decision-making in fields like finance and
healthcare. Their importance has driven a recent influx of works passing time
series into language models, leading to non-trivial forecasting on some
datasets. But it remains unknown whether non-trivial forecasting implies that
language models can reason about time series. To address this gap, we generate
a first-of-its-kind evaluation framework for time series reasoning, including
formal tasks and a corresponding dataset of multi-scale time series paired with
text captions across ten domains. Using these data, we probe whether language
models achieve three forms of reasoning: (1) Etiological Reasoning - given an
input time series, can the language model identify the scenario that most
likely created it? (2) Question Answering - can a language model answer factual
questions about time series? (3) Context-Aided Forecasting - does highly
relevant textual context improve a language model's time series forecasts?
We find that otherwise highly-capable language models demonstrate
surprisingly limited time series reasoning: they score marginally above random
on etiological and question answering tasks (up to 30 percentage points worse
than humans) and show modest success in using context to improve forecasting.
These weakness showcase that time series reasoning is an impactful, yet deeply
underdeveloped direction for language model research. We also make our datasets
and code public at to support further research in this direction at
https://github.com/behavioral-data/TSandLanguage | cs.CL | Language Models Still Struggle to Zero-shot Reason about Time Series |
2024-04-17T21:09:13Z | http://arxiv.org/pdf/2404.11752v1 | http://arxiv.org/abs/2404.11752v1 | Nazia Tasnim, Sujan Sen Gupta, Md. Istiak Hossain Shihab, Fatiha Islam Juee, Arunima Tahsin, Pritom Ghum, Kanij Fatema, Marshia Haque, Wasema Farzana, Prionti Nasir, Ashique KhudaBukhsh, Farig Sadeque, Asif Sushmit | 2024-04-17T21:09:13Z | cs.CL, cs.CY | Communal violence in online forums has become extremely prevalent in South
Asia, where many communities of different cultures coexist and share resources.
These societies exhibit a phenomenon characterized by strong bonds within their
own groups and animosity towards others, leading to conflicts that frequently
escalate into violent confrontations. To address this issue, we have developed
the first comprehensive framework for the automatic detection of communal
violence markers in online Bangla content accompanying the largest collection
(13K raw sentences) of social media interactions that fall under the definition
of four major violence class and their 16 coarse expressions. Our workflow
introduces a 7-step expert annotation process incorporating insights from
social scientists, linguists, and psychologists. By presenting data statistics
and benchmarking performance using this dataset, we have determined that, aside
from the category of Non-communal violence, Religio-communal violence is
particularly pervasive in Bangla text. Moreover, we have substantiated the
effectiveness of fine-tuning language models in identifying violent comments by
conducting preliminary benchmarking on the state-of-the-art Bangla deep
learning model. | cs.CL | Mapping Violence: Developing an Extensive Framework to Build a Bangla Sectarian Expression Dataset from Social Media Interactions |
2024-04-17T20:31:05Z | http://arxiv.org/pdf/2404.11730v2 | http://arxiv.org/abs/2404.11730v2 | Graham Todd, Tim Merino, Sam Earle, Julian Togelius | 2024-04-21T15:38:19Z | cs.CL, cs.AI | The Connections puzzle published each day by the New York Times tasks players
with dividing a bank of sixteen words into four groups of four words that each
relate to a common theme. Solving the puzzle requires both common linguistic
knowledge (i.e. definitions and typical usage) as well as, in many cases,
lateral or abstract thinking. This is because the four categories ascend in
complexity, with the most challenging category often requiring thinking about
words in uncommon ways or as parts of larger phrases. We investigate the
capacity for automated AI systems to play Connections and explore the game's
potential as an automated benchmark for abstract reasoning and a way to measure
the semantic information encoded by data-driven linguistic systems. In
particular, we study both a sentence-embedding baseline and modern large
language models (LLMs). We report their accuracy on the task, measure the
impacts of chain-of-thought prompting, and discuss their failure modes.
Overall, we find that the Connections task is challenging yet feasible, and a
strong test-bed for future work. | cs.CL | Missed Connections: Lateral Thinking Puzzles for Large Language Models |
2024-04-17T20:24:41Z | http://arxiv.org/pdf/2404.11726v1 | http://arxiv.org/abs/2404.11726v1 | Orhun Caglidil, Malte Ostendorff, Georg Rehm | 2024-04-17T20:24:41Z | cs.CL | Language models are trained mostly on Web data, which often contains social
stereotypes and biases that the models can inherit. This has potentially
negative consequences, as models can amplify these biases in downstream tasks
or applications. However, prior research has primarily focused on the English
language, especially in the context of gender bias. In particular,
grammatically gender-neutral languages such as Turkish are underexplored
despite representing different linguistic properties to language models with
possibly different effects on biases. In this paper, we fill this research gap
and investigate the significance of gender bias in Turkish language models. We
build upon existing bias evaluation frameworks and extend them to the Turkish
language by translating existing English tests and creating new ones designed
to measure gender bias in the context of T\"urkiye. Specifically, we also
evaluate Turkish language models for their embedded ethnic bias toward Kurdish
people. Based on the experimental results, we attribute possible biases to
different model characteristics such as the model size, their multilingualism,
and the training corpora. We make the Turkish gender bias dataset publicly
available. | cs.CL | Investigating Gender Bias in Turkish Language Models |
2024-04-17T20:11:32Z | http://arxiv.org/pdf/2404.11717v1 | http://arxiv.org/abs/2404.11717v1 | Neha Srikanth, Marine Carpuat, Rachel Rudinger | 2024-04-17T20:11:32Z | cs.CL | Large language models have been shown to behave inconsistently in response to
meaning-preserving paraphrastic inputs. At the same time, researchers evaluate
the knowledge and reasoning abilities of these models with test evaluations
that do not disaggregate the effect of paraphrastic variability on performance.
We propose a metric for evaluating the paraphrastic consistency of natural
language reasoning models based on the probability of a model achieving the
same correctness on two paraphrases of the same problem. We mathematically
connect this metric to the proportion of a model's variance in correctness
attributable to paraphrasing. To estimate paraphrastic consistency, we collect
ParaNLU, a dataset of 7,782 human-written and validated paraphrased reasoning
problems constructed on top of existing benchmark datasets for defeasible and
abductive natural language inference. Using ParaNLU, we measure the
paraphrastic consistency of several model classes and show that consistency
dramatically increases with pretraining but not finetuning. All models tested
exhibited room for improvement in paraphrastic consistency. | cs.CL | How often are errors in natural language reasoning due to paraphrastic variability? |
2024-04-17T19:53:59Z | http://arxiv.org/pdf/2404.13087v1 | http://arxiv.org/abs/2404.13087v1 | Shikha Soneji, Mitchell Hoesing, Sujay Koujalgi, Jonathan Dodge | 2024-04-17T19:53:59Z | cs.CL, cs.LG | The complexities of legalese in terms and policy documents can bind
individuals to contracts they do not fully comprehend, potentially leading to
uninformed data sharing. Our work seeks to alleviate this issue by developing
language models that provide automated, accessible summaries and scores for
such documents, aiming to enhance user understanding and facilitate informed
decisions. We compared transformer-based and conventional models during
training on our dataset, and RoBERTa performed better overall with a remarkable
0.74 F1-score. Leveraging our best-performing model, RoBERTa, we highlighted
redundancies and potential guideline violations by identifying overlaps in
GDPR-required documents, underscoring the necessity for stricter GDPR
compliance. | cs.CL | Demystifying Legalese: An Automated Approach for Summarizing and Analyzing Overlaps in Privacy Policies and Terms of Service |
2024-04-17T18:42:36Z | http://arxiv.org/pdf/2404.11691v1 | http://arxiv.org/abs/2404.11691v1 | Vansh Gupta, Mohit Gupta, Jai Garg, Nitesh Garg | 2024-04-17T18:42:36Z | cs.CL | Address matching is an important task for many businesses especially delivery
and take out companies which help them to take out a certain address from their
data warehouse. Existing solution uses similarity of strings, and edit distance
algorithms to find out the similar addresses from the address database, but
these algorithms could not work effectively with redundant, unstructured, or
incomplete address data. This paper discuss semantic Address matching
technique, by which we can find out a particular address from a list of
possible addresses. We have also reviewed existing practices and their
shortcoming. Semantic address matching is an essentially NLP task in the field
of deep learning. Through this technique We have the ability to triumph the
drawbacks of existing methods like redundant or abbreviated data problems. The
solution uses the OCR on invoices to extract the address and create the data
pool of addresses. Then this data is fed to the algorithm BM-25 for scoring the
best matching entries. Then to observe the best result, this will pass through
BERT for giving the best possible result from the similar queries. Our
investigation exhibits that our methodology enormously improves both accuracy
and review of cutting-edge technology existing techniques. | cs.CL | Improvement in Semantic Address Matching using Natural Language Processing |
2024-04-17T18:27:59Z | http://arxiv.org/pdf/2404.11682v1 | http://arxiv.org/abs/2404.11682v1 | Mahsa Sheikhi Karizaki, Dana Gnesdilow, Sadhana Puntambekar, Rebecca J. Passonneau | 2024-04-17T18:27:59Z | cs.CL | Automated methods are becoming increasingly integrated into studies of
formative feedback on students' science explanation writing. Most of this work,
however, addresses students' responses to short answer questions. We
investigate automated feedback on students' science explanation essays, where
students must articulate multiple ideas. Feedback is based on a rubric that
identifies the main ideas students are prompted to include in explanatory
essays about the physics of energy and mass, given their experiments with a
simulated roller coaster. We have found that students generally improve on
revised versions of their essays. Here, however, we focus on two factors that
affect the accuracy of the automated feedback. First, we find that the main
ideas in the rubric differ with respect to how much freedom they afford in
explanations of the idea, thus explanation of a natural law is relatively
constrained. Students have more freedom in how they explain complex relations
they observe in their roller coasters, such as transfer of different forms of
energy. Second, by tracing the automated decision process, we can diagnose when
a student's statement lacks sufficient clarity for the automated tool to
associate it more strongly with one of the main ideas above all others. This in
turn provides an opportunity for teachers and peers to help students reflect on
how to state their ideas more clearly. | cs.CL | How Well Can You Articulate that Idea? Insights from Automated Formative Assessment |
2024-04-17T18:13:16Z | http://arxiv.org/pdf/2404.11672v1 | http://arxiv.org/abs/2404.11672v1 | Ali Modarressi, Abdullatif Köksal, Ayyoob Imani, Mohsen Fayyaz, Hinrich Schütze | 2024-04-17T18:13:16Z | cs.CL | While current large language models (LLMs) demonstrate some capabilities in
knowledge-intensive tasks, they are limited by relying on their parameters as
an implicit storage mechanism. As a result, they struggle with infrequent
knowledge and temporal degradation. In addition, the uninterpretable nature of
parametric memorization makes it challenging to understand and prevent
hallucination. Parametric memory pools and model editing are only partial
solutions. Retrieval Augmented Generation (RAG) $\unicode{x2013}$ though
non-parametric $\unicode{x2013}$ has its own limitations: it lacks structure,
complicates interpretability and makes it hard to effectively manage stored
knowledge. In this paper, we introduce MemLLM, a novel method of enhancing LLMs
by integrating a structured and explicit read-and-write memory module. MemLLM
tackles the aforementioned challenges by enabling dynamic interaction with the
memory and improving the LLM's capabilities in using stored knowledge. Our
experiments indicate that MemLLM enhances the LLM's performance and
interpretability, in language modeling in general and knowledge-intensive tasks
in particular. We see MemLLM as an important step towards making LLMs more
grounded and factual through memory augmentation. | cs.CL | MemLLM: Finetuning LLMs to Use An Explicit Read-Write Memory |
2024-04-17T17:37:30Z | http://arxiv.org/pdf/2404.11588v1 | http://arxiv.org/abs/2404.11588v1 | Xiangci Li, Jessica Ouyang | 2024-04-17T17:37:30Z | cs.CL | To convince readers of the novelty of their research paper, authors must
perform a literature review and compose a coherent story that connects and
relates prior works to the current work. This challenging nature of literature
review writing makes automatic related work generation (RWG) academically and
computationally interesting, and also makes it an excellent test bed for
examining the capability of SOTA natural language processing (NLP) models.
Since the initial proposal of the RWG task, its popularity has waxed and waned,
following the capabilities of mainstream NLP approaches. In this work, we
survey the zoo of RWG historical works, summarizing the key approaches and task
definitions and discussing the ongoing challenges of RWG. | cs.CL | Related Work and Citation Text Generation: A Survey |
2024-04-17T17:32:41Z | http://arxiv.org/pdf/2404.11584v1 | http://arxiv.org/abs/2404.11584v1 | Tula Masterman, Sandi Besen, Mason Sawtell, Alex Chao | 2024-04-17T17:32:41Z | cs.AI, cs.CL | This survey paper examines the recent advancements in AI agent
implementations, with a focus on their ability to achieve complex goals that
require enhanced reasoning, planning, and tool execution capabilities. The
primary objectives of this work are to a) communicate the current capabilities
and limitations of existing AI agent implementations, b) share insights gained
from our observations of these systems in action, and c) suggest important
considerations for future developments in AI agent design. We achieve this by
providing overviews of single-agent and multi-agent architectures, identifying
key patterns and divergences in design choices, and evaluating their overall
impact on accomplishing a provided goal. Our contribution outlines key themes
when selecting an agentic architecture, the impact of leadership on agent
systems, agent communication styles, and key phases for planning, execution,
and reflection that enable robust AI agent systems. | cs.AI | The Landscape of Emerging AI Agent Architectures for Reasoning, Planning, and Tool Calling: A Survey |
2024-04-17T16:53:16Z | http://arxiv.org/pdf/2404.11553v2 | http://arxiv.org/abs/2404.11553v2 | Zihao Li, Yucheng Shi, Zirui Liu, Fan Yang, Ali Payani, Ninghao Liu, Mengnan Du | 2024-06-16T08:24:32Z | cs.CL, cs.AI, cs.LG | The development of Large Language Models (LLMs) relies on extensive text
corpora, which are often unevenly distributed across languages. This imbalance
results in LLMs performing significantly better on high-resource languages like
English, German, and French, while their capabilities in low-resource languages
remain inadequate. Currently, there is a lack of quantitative methods to
evaluate the performance of LLMs in these low-resource languages. To address
this gap, we propose the Language Ranker, an intrinsic metric designed to
benchmark and rank languages based on LLM performance using internal
representations. By comparing the LLM's internal representation of various
languages against a baseline derived from English, we can assess the model's
multilingual capabilities in a robust and language-agnostic manner. Our
analysis reveals that high-resource languages exhibit higher similarity scores
with English, demonstrating superior performance, while low-resource languages
show lower similarity scores, underscoring the effectiveness of our metric in
assessing language-specific capabilities. Besides, the experiments show that
there is a strong correlation between the LLM's performance in different
languages and the proportion of those languages in its pre-training corpus.
These insights underscore the efficacy of the Language Ranker as a tool for
evaluating LLM performance across different languages, particularly those with
limited resources. | cs.CL | Quantifying Multilingual Performance of Large Language Models Across Languages |
2024-04-17T16:33:22Z | http://arxiv.org/pdf/2404.11539v1 | http://arxiv.org/abs/2404.11539v1 | Soyoung Yang, Won Ik Cho | 2024-04-17T16:33:22Z | cs.CL | In the era of rapid evolution of generative language models within the realm
of natural language processing, there is an imperative call to revisit and
reformulate evaluation methodologies, especially in the domain of aspect-based
sentiment analysis (ABSA). This paper addresses the emerging challenges
introduced by the generative paradigm, which has moderately blurred traditional
boundaries between understanding and generation tasks. Building upon prevailing
practices in the field, we analyze the advantages and shortcomings associated
with the prevalent ABSA evaluation paradigms. Through an in-depth examination,
supplemented by illustrative examples, we highlight the intricacies involved in
aligning generative outputs with other evaluative metrics, specifically those
derived from other tasks, including question answering. While we steer clear of
advocating for a singular and definitive metric, our contribution lies in
paving the path for a comprehensive guideline tailored for ABSA evaluations in
this generative paradigm. In this position paper, we aim to provide
practitioners with profound reflections, offering insights and directions that
can aid in navigating this evolving landscape, ensuring evaluations that are
both accurate and reflective of generative capabilities. | cs.CL | Evaluating Span Extraction in Generative Paradigm: A Reflection on Aspect-Based Sentiment Analysis |
2024-04-17T16:32:13Z | http://arxiv.org/pdf/2404.11538v1 | http://arxiv.org/abs/2404.11538v1 | Md Athikul Islam, Edoardo Serra, Sushil Jajodia | 2024-04-17T16:32:13Z | cs.LG, cs.CL | Adversarial attacks pose significant challenges to deep neural networks
(DNNs) such as Transformer models in natural language processing (NLP). This
paper introduces a novel defense strategy, called GenFighter, which enhances
adversarial robustness by learning and reasoning on the training classification
distribution. GenFighter identifies potentially malicious instances deviating
from the distribution, transforms them into semantically equivalent instances
aligned with the training data, and employs ensemble techniques for a unified
and robust response. By conducting extensive experiments, we show that
GenFighter outperforms state-of-the-art defenses in accuracy under attack and
attack success rate metrics. Additionally, it requires a high number of queries
per attack, making the attack more challenging in real scenarios. The ablation
study shows that our approach integrates transfer learning, a
generative/evolutive procedure, and an ensemble method, providing an effective
defense against NLP adversarial attacks. | cs.LG | GenFighter: A Generative and Evolutive Textual Attack Removal |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.