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
|
|---|---|---|---|---|---|---|---|---|
2023-11-09T08:00:32Z | http://arxiv.org/pdf/2311.05661v3 | http://arxiv.org/abs/2311.05661v3 | Qinyuan Ye, Maxamed Axmed, Reid Pryzant, Fereshte Khani | 2024-07-03T01:29:20Z | cs.CL, cs.AI, cs.LG | Prompt engineering is a challenging yet crucial task for optimizing the
performance of large language models on customized tasks. It requires complex
reasoning to examine the model's errors, hypothesize what is missing or
misleading in the current prompt, and communicate the task with clarity. While
recent works indicate that large language models can be meta-prompted to
perform automatic prompt engineering, we argue that their potential is limited
due to insufficient guidance for complex reasoning in the meta-prompt. We fill
this gap by infusing into the meta-prompt three key components: detailed
descriptions, context specification, and a step-by-step reasoning template. The
resulting method, named PE2, exhibits remarkable versatility across diverse
language tasks. It finds prompts that outperform "let's think step by step" by
6.3% on MultiArith and 3.1% on GSM8K, and outperforms competitive baselines on
counterfactual tasks by 6.9%. Further, we show that PE2 can make targeted and
highly specific prompt edits, rectify erroneous prompts, and induce multi-step
plans for complex tasks. | cs.CL | Prompt Engineering a Prompt Engineer |
2023-11-09T06:45:04Z | http://arxiv.org/pdf/2311.05169v1 | http://arxiv.org/abs/2311.05169v1 | Diego Mollá | 2023-11-09T06:45:04Z | cs.CL | This paper reports on the use of prompt engineering and GPT-3.5 for
biomedical query-focused multi-document summarisation. Using GPT-3.5 and
appropriate prompts, our system achieves top ROUGE-F1 results in the task of
obtaining short-paragraph-sized answers to biomedical questions in the 2023
BioASQ Challenge (BioASQ 11b). This paper confirms what has been observed in
other domains: 1) Prompts that incorporated few-shot samples generally improved
on their counterpart zero-shot variants; 2) The largest improvement was
achieved by retrieval augmented generation. The fact that these prompts allow
our top runs to rank within the top two runs of BioASQ 11b demonstrate the
power of using adequate prompts for Large Language Models in general, and
GPT-3.5 in particular, for query-focused summarisation. | cs.CL | Large Language Models and Prompt Engineering for Biomedical Query Focused Multi-Document Summarisation |
2023-11-09T06:19:51Z | http://arxiv.org/pdf/2311.05161v1 | http://arxiv.org/abs/2311.05161v1 | Jangwhan Lee, Minsoo Kim, Seungcheol Baek, Seok Joong Hwang, Wonyong Sung, Jungwook Choi | 2023-11-09T06:19:51Z | cs.CL | Large Language Models (LLMs) are proficient in natural language processing
tasks, but their deployment is often restricted by extensive parameter sizes
and computational demands. This paper focuses on post-training quantization
(PTQ) in LLMs, specifically 4-bit weight and 8-bit activation (W4A8)
quantization, to enhance computational efficiency -- a topic less explored
compared to weight-only quantization. We present two innovative techniques:
activation-quantization-aware scaling (AQAS) and sequence-length-aware
calibration (SLAC) to enhance PTQ by considering the combined effects on
weights and activations and aligning calibration sequence lengths to target
tasks. Moreover, we introduce dINT, a hybrid data format combining integer and
denormal representations, to address the underflow issue in W4A8 quantization,
where small values are rounded to zero. Through rigorous evaluations of LLMs,
including OPT and LLaMA, we demonstrate that our techniques significantly boost
task accuracies to levels comparable with full-precision models. By developing
arithmetic units compatible with dINT, we further confirm that our methods
yield a 2$\times$ hardware efficiency improvement compared to 8-bit integer MAC
unit. | cs.CL | Enhancing Computation Efficiency in Large Language Models through Weight and Activation Quantization |
2023-11-09T06:03:11Z | http://arxiv.org/pdf/2311.07589v1 | http://arxiv.org/abs/2311.07589v1 | Yerin Hwang, Yongil Kim, Hyunkyung Bae, Jeesoo Bang, Hwanhee Lee, Kyomin Jung | 2023-11-09T06:03:11Z | cs.CL, cs.AI | To address the data scarcity issue in Conversational question answering
(ConvQA), a dialog inpainting method, which utilizes documents to generate
ConvQA datasets, has been proposed. However, the original dialog inpainting
model is trained solely on the dialog reconstruction task, resulting in the
generation of questions with low contextual relevance due to insufficient
learning of question-answer alignment. To overcome this limitation, we propose
a novel framework called Dialogizer, which has the capability to automatically
generate ConvQA datasets with high contextual relevance from textual sources.
The framework incorporates two training tasks: question-answer matching (QAM)
and topic-aware dialog generation (TDG). Moreover, re-ranking is conducted
during the inference phase based on the contextual relevance of the generated
questions. Using our framework, we produce four ConvQA datasets by utilizing
documents from multiple domains as the primary source. Through automatic
evaluation using diverse metrics, as well as human evaluation, we validate that
our proposed framework exhibits the ability to generate datasets of higher
quality compared to the baseline dialog inpainting model. | cs.CL | Dialogizer: Context-aware Conversational-QA Dataset Generation from Textual Sources |
2023-11-09T05:46:41Z | http://arxiv.org/pdf/2311.05155v1 | http://arxiv.org/abs/2311.05155v1 | Koustava Goswami, Priya Rani, Theodorus Fransen, John P. McCrae | 2023-11-09T05:46:41Z | cs.CL, cs.AI | Exploiting cognates for transfer learning in under-resourced languages is an
exciting opportunity for language understanding tasks, including unsupervised
machine translation, named entity recognition and information retrieval.
Previous approaches mainly focused on supervised cognate detection tasks based
on orthographic, phonetic or state-of-the-art contextual language models, which
under-perform for most under-resourced languages. This paper proposes a novel
language-agnostic weakly-supervised deep cognate detection framework for
under-resourced languages using morphological knowledge from closely related
languages. We train an encoder to gain morphological knowledge of a language
and transfer the knowledge to perform unsupervised and weakly-supervised
cognate detection tasks with and without the pivot language for the
closely-related languages. While unsupervised, it overcomes the need for
hand-crafted annotation of cognates. We performed experiments on different
published cognate detection datasets across language families and observed not
only significant improvement over the state-of-the-art but also our method
outperformed the state-of-the-art supervised and unsupervised methods. Our
model can be extended to a wide range of languages from any language family as
it overcomes the requirement of the annotation of the cognate pairs for
training. The code and dataset building scripts can be found at
https://github.com/koustavagoswami/Weakly_supervised-Cognate_Detection | cs.CL | Weakly-supervised Deep Cognate Detection Framework for Low-Resourced Languages Using Morphological Knowledge of Closely-Related Languages |
2023-11-09T03:14:54Z | http://arxiv.org/pdf/2311.05120v1 | http://arxiv.org/abs/2311.05120v1 | Yasser Shohoud, Maged Shoman, Sarah Abdelazim | 2023-11-09T03:14:54Z | cs.CL | The Holy Book of Quran is believed to be the literal word of God (Allah) as
revealed to the Prophet Muhammad (PBUH) over a period of approximately 23
years. It is the book where God provides guidance on how to live a righteous
and just life, emphasizing principles like honesty, compassion, charity and
justice, as well as providing rules for personal conduct, family matters,
business ethics and much more. However, due to constraints related to the
language and the Quran organization, it is challenging for Muslims to get all
relevant ayahs (verses) pertaining to a matter or inquiry of interest. Hence,
we developed a Quran semantic search tool which finds the verses pertaining to
the user inquiry or prompt. To achieve this, we trained several models on a
large dataset of over 30 tafsirs, where typically each tafsir corresponds to
one verse in the Quran and, using cosine similarity, obtained the tafsir tensor
which is most similar to the prompt tensor of interest, which was then used to
index for the corresponding ayah in the Quran. Using the SNxLM model, we were
able to achieve a cosine similarity score as high as 0.97 which corresponds to
the abdu tafsir for a verse relating to financial matters. | cs.CL | Quranic Conversations: Developing a Semantic Search tool for the Quran using Arabic NLP Techniques |
2023-11-09T03:10:42Z | http://arxiv.org/pdf/2311.05117v1 | http://arxiv.org/abs/2311.05117v1 | Yuto Kuroda, Atsushi Fujita, Tomoyuki Kajiwara, Takashi Ninomiya | 2023-11-09T03:10:42Z | cs.CL | Translation quality estimation (TQE) is the task of predicting translation
quality without reference translations. Due to the enormous cost of creating
training data for TQE, only a few translation directions can benefit from
supervised training. To address this issue, unsupervised TQE methods have been
studied. In this paper, we extensively investigate the usefulness of synthetic
TQE data and pre-trained multilingual encoders in unsupervised sentence-level
TQE, both of which have been proven effective in the supervised training
scenarios. Our experiment on WMT20 and WMT21 datasets revealed that this
approach can outperform other unsupervised TQE methods on high- and
low-resource translation directions in predicting post-editing effort and human
evaluation score, and some zero-resource translation directions in predicting
post-editing effort. | cs.CL | Unsupervised Translation Quality Estimation Exploiting Synthetic Data and Pre-trained Multilingual Encoder |
2023-11-09T02:58:17Z | http://arxiv.org/pdf/2311.05113v1 | http://arxiv.org/abs/2311.05113v1 | Haoyi Wu, Wenyang Hui, Yezeng Chen, Weiqi Wu, Kewei Tu, Yi Zhou | 2023-11-09T02:58:17Z | cs.CL | Mathematical understanding and reasoning are crucial tasks for assessing the
capabilities of artificial intelligence (AI). However, existing benchmarks
either require just a few steps of reasoning, or only contain a small amount of
data in one specific topic, making it hard to analyse AI's behaviour with
reference to different problems within a specific topic in detail. In this
work, we propose Conic10K, a challenging math problem dataset on conic sections
in Chinese senior high school education. Our dataset contains various problems
with different reasoning depths, while only the knowledge from conic sections
is required. Since the dataset only involves a narrow range of knowledge, it is
easy to separately analyse the knowledge a model possesses and the reasoning
ability it has. For each problem, we provide a high-quality formal
representation, the reasoning steps, and the final solution. Experiments show
that existing large language models, including GPT-4, exhibit weak performance
on complex reasoning. We hope that our findings could inspire more advanced
techniques for precise natural language understanding and reasoning. Our
dataset and codes are available at https://github.com/whyNLP/Conic10K. | cs.CL | Conic10K: A Challenging Math Problem Understanding and Reasoning Dataset |
2023-11-09T02:55:58Z | http://arxiv.org/pdf/2311.05112v5 | http://arxiv.org/abs/2311.05112v5 | Hongjian Zhou, Fenglin Liu, Boyang Gu, Xinyu Zou, Jinfa Huang, Jinge Wu, Yiru Li, Sam S. Chen, Peilin Zhou, Junling Liu, Yining Hua, Chengfeng Mao, Chenyu You, Xian Wu, Yefeng Zheng, Lei Clifton, Zheng Li, Jiebo Luo, David A. Clifton | 2024-05-15T13:38:45Z | cs.CL, cs.AI | Large language models (LLMs), such as ChatGPT, have received substantial
attention due to their capabilities for understanding and generating human
language. While there has been a burgeoning trend in research focusing on the
employment of LLMs in supporting different medical tasks (e.g., enhancing
clinical diagnostics and providing medical education), a review of these
efforts, particularly their development, practical applications, and outcomes
in medicine, remains scarce. Therefore, this review aims to provide a detailed
overview of the development and deployment of LLMs in medicine, including the
challenges and opportunities they face. In terms of development, we provide a
detailed introduction to the principles of existing medical LLMs, including
their basic model structures, number of parameters, and sources and scales of
data used for model development. It serves as a guide for practitioners in
developing medical LLMs tailored to their specific needs. In terms of
deployment, we offer a comparison of the performance of different LLMs across
various medical tasks, and further compare them with state-of-the-art
lightweight models, aiming to provide an understanding of the advantages and
limitations of LLMs in medicine. Overall, in this review, we address the
following questions: 1) What are the practices for developing medical LLMs 2)
How to measure the medical task performance of LLMs in a medical setting? 3)
How have medical LLMs been employed in real-world practice? 4) What challenges
arise from the use of medical LLMs? and 5) How to more effectively develop and
deploy medical LLMs? By answering these questions, this review aims to provide
insights into the opportunities for LLMs in medicine and serve as a practical
resource. We also maintain a regularly updated list of practical guides on
medical LLMs at: https://github.com/AI-in-Health/MedLLMsPracticalGuide. | cs.CL | A Survey of Large Language Models in Medicine: Progress, Application, and Challenge |
2023-11-09T01:27:54Z | http://arxiv.org/pdf/2311.05089v1 | http://arxiv.org/abs/2311.05089v1 | Daniele Giofré, Sneha Ghantasala | 2023-11-09T01:27:54Z | cs.CL, cs.AI | Since its introduction, the transformers architecture has seen great adoption
in NLP applications, but it also has limitations. Although the self-attention
mechanism allows for generating very rich representations of the input text,
its effectiveness may be limited in specialized domains such as legal, where,
for example, language models often have to process very long texts. In this
paper, we explore alternatives to replace the attention-based layers with
simpler token-mixing mechanisms: Hartley and Fourier transforms. Using these
non-parametric techniques, we train models with long input documents from
scratch in the legal domain setting. We also introduce a new hybrid Seq2Seq
architecture, a no-attention-based encoder connected with an attention-based
decoder, which performs quite well on existing summarization tasks with much
less compute and memory requirements. We believe that similar, if not better
performance, as in the case of long correlations of abstractive text
summarization tasks, can be achieved by adopting these simpler infrastructures.
This not only makes training models from scratch accessible to more people, but
also contributes to the reduction of the carbon footprint during training. | cs.CL | Legal-HNet: Mixing Legal Long-Context Tokens with Hartley Transform |
2023-11-09T01:04:44Z | http://arxiv.org/pdf/2311.05085v2 | http://arxiv.org/abs/2311.05085v2 | Aditi Mishra, Sajjadur Rahman, Hannah Kim, Kushan Mitra, Estevam Hruschka | 2024-01-31T19:17:00Z | cs.CL, cs.AI | Large language models (LLMs) are proficient at generating fluent text with
minimal task-specific supervision. Yet, their ability to provide well-grounded
rationalizations for knowledge-intensive tasks remains under-explored. Such
tasks, like commonsense multiple-choice questions, require rationales based on
world knowledge to support predictions and refute alternate options. We
consider the task of generating knowledge-guided rationalization in natural
language by using expert-written examples in a few-shot manner. Surprisingly,
crowd-workers preferred knowledge-grounded rationales over crowdsourced
rationalizations, citing their factuality, sufficiency, and comprehensive
refutations. Although LLMs-generated rationales were preferable, further
improvements in conciseness and novelty are required. In another study, we show
how rationalization of incorrect model predictions erodes humans' trust in
LLM-generated rationales. Motivated by these observations, we create a
two-stage pipeline to review task predictions and eliminate potential incorrect
decisions before rationalization, enabling trustworthy rationale generation. | cs.CL | Characterizing Large Language Models as Rationalizers of Knowledge-intensive Tasks |
2023-11-09T00:30:13Z | http://arxiv.org/pdf/2311.05657v2 | http://arxiv.org/abs/2311.05657v2 | Da Yin, Faeze Brahman, Abhilasha Ravichander, Khyathi Chandu, Kai-Wei Chang, Yejin Choi, Bill Yuchen Lin | 2024-03-13T10:54:21Z | cs.AI, cs.CL, cs.LG | Closed-source agents suffer from several issues such as a lack of
affordability, transparency, and reproducibility, particularly on complex
interactive tasks. This motivates the development of open-source alternatives.
We introduce LUMOS, one of the first frameworks for training open-source
LLM-based agents. LUMOS features a learnable, unified, and modular architecture
with a planning module that learns high-level subgoal generation, and a
grounding module trained to translate these into actions using various tools in
the execution module. The design allows for modular upgrades and wider
applicability to diverse interactive tasks. To foster generalizable agent
learning, we collect large-scale, unified, and high-quality training
annotations derived from diverse ground-truth reasoning rationales across
various complex interactive tasks. On 9 datasets, LUMOS exhibits several key
advantages: (1) LUMOS excels multiple larger open-source agents on the held-out
datasets (unused for training) for each task type. LUMOS even surpasses GPT
agents on QA and web tasks; (2) LUMOS outperforms open-source agents produced
by chain-of-thoughts and unmodularized integrated training; and (3) LUMOS
effectively generalizes to unseen tasks, outperforming 33B-scale agents and
domain-specific agents. | cs.AI | Agent Lumos: Unified and Modular Training for Open-Source Language Agents |
2023-11-09T00:27:28Z | http://arxiv.org/pdf/2312.07751v3 | http://arxiv.org/abs/2312.07751v3 | Nikita Soni, H. Andrew Schwartz, João Sedoc, Niranjan Balasubramanian | 2024-05-09T17:22:40Z | cs.CL, cs.AI, cs.LG | As research in human-centered NLP advances, there is a growing recognition of
the importance of incorporating human and social factors into NLP models. At
the same time, our NLP systems have become heavily reliant on LLMs, most of
which do not model authors. To build NLP systems that can truly understand
human language, we must better integrate human contexts into LLMs. This brings
to the fore a range of design considerations and challenges in terms of what
human aspects to capture, how to represent them, and what modeling strategies
to pursue. To address these, we advocate for three positions toward creating
large human language models (LHLMs) using concepts from psychological and
behavioral sciences: First, LM training should include the human context.
Second, LHLMs should recognize that people are more than their group(s). Third,
LHLMs should be able to account for the dynamic and temporally-dependent nature
of the human context. We refer to relevant advances and present open challenges
that need to be addressed and their possible solutions in realizing these
goals. | cs.CL | Large Human Language Models: A Need and the Challenges |
2023-11-09T00:15:06Z | http://arxiv.org/pdf/2311.05075v1 | http://arxiv.org/abs/2311.05075v1 | Haijian Shao, Ming Zhu, Shengjie Zhai | 2023-11-09T00:15:06Z | cs.LG, cs.AI, cs.CL | Amid growing global mental health concerns, particularly among vulnerable
groups, natural language processing offers a tremendous potential for early
detection and intervention of people's mental disorders via analyzing their
postings and discussions on social media platforms. However, ultra-sparse
training data, often due to vast vocabularies and low-frequency words, hinders
the analysis accuracy. Multi-labeling and Co-occurrences of symptoms may also
blur the boundaries in distinguishing similar/co-related disorders. To address
these issues, we propose a novel semantic feature preprocessing technique with
a three-folded structure: 1) mitigating the feature sparsity with a weak
classifier, 2) adaptive feature dimension with modulus loops, and 3)
deep-mining and extending features among the contexts. With enhanced semantic
features, we train a machine learning model to predict and classify mental
disorders. We utilize the Reddit Mental Health Dataset 2022 to examine
conditions such as Anxiety, Borderline Personality Disorder (BPD), and
Bipolar-Disorder (BD) and present solutions to the data sparsity challenge,
highlighted by 99.81% non-zero elements. After applying our preprocessing
technique, the feature sparsity decreases to 85.4%. Overall, our methods, when
compared to seven benchmark models, demonstrate significant performance
improvements: 8.0% in accuracy, 0.069 in precision, 0.093 in recall, 0.102 in
F1 score, and 0.059 in AUC. This research provides foundational insights for
mental health prediction and monitoring, providing innovative solutions to
navigate challenges associated with ultra-sparse data feature and intricate
multi-label classification in the domain of mental health analysis. | cs.LG | Mental Health Diagnosis in the Digital Age: Harnessing Sentiment Analysis on Social Media Platforms upon Ultra-Sparse Feature Content |
2023-11-09T00:12:21Z | http://arxiv.org/pdf/2311.05074v2 | http://arxiv.org/abs/2311.05074v2 | Vinodkumar Prabhakaran, Christopher Homan, Lora Aroyo, Aida Mostafazadeh Davani, Alicia Parrish, Alex Taylor, Mark Díaz, Ding Wang, Gregory Serapio-García | 2024-06-14T03:28:49Z | cs.CL, cs.AI | Human annotation plays a core role in machine learning -- annotations for
supervised models, safety guardrails for generative models, and human feedback
for reinforcement learning, to cite a few avenues. However, the fact that many
of these human annotations are inherently subjective is often overlooked.
Recent work has demonstrated that ignoring rater subjectivity (typically
resulting in rater disagreement) is problematic within specific tasks and for
specific subgroups. Generalizable methods to harness rater disagreement and
thus understand the socio-cultural leanings of subjective tasks remain elusive.
In this paper, we propose GRASP, a comprehensive disagreement analysis
framework to measure group association in perspectives among different rater
sub-groups, and demonstrate its utility in assessing the extent of systematic
disagreements in two datasets: (1) safety annotations of human-chatbot
conversations, and (2) offensiveness annotations of social media posts, both
annotated by diverse rater pools across different socio-demographic axes. Our
framework (based on disagreement metrics) reveals specific rater groups that
have significantly different perspectives than others on certain tasks, and
helps identify demographic axes that are crucial to consider in specific task
contexts. | cs.CL | GRASP: A Disagreement Analysis Framework to Assess Group Associations in Perspectives |
2023-11-08T22:58:34Z | http://arxiv.org/pdf/2311.05051v1 | http://arxiv.org/abs/2311.05051v1 | Juliana Resplande Santanna Gomes, Eduardo Augusto Santos Garcia, Adalberto Ferreira Barbosa Junior, Ruan Chaves Rodrigues, Diogo Fernandes Costa Silva, Dyonnatan Ferreira Maia, Nádia Félix Felipe da Silva, Arlindo Rodrigues Galvão Filho, Anderson da Silva Soares | 2023-11-08T22:58:34Z | cs.CL | Aspect-based Sentiment Analysis (ABSA) is a task whose objective is to
classify the individual sentiment polarity of all entities, called aspects, in
a sentence. The task is composed of two subtasks: Aspect Term Extraction (ATE),
identify all aspect terms in a sentence; and Sentiment Orientation Extraction
(SOE), given a sentence and its aspect terms, the task is to determine the
sentiment polarity of each aspect term (positive, negative or neutral). This
article presents we present our participation in Aspect-Based Sentiment
Analysis in Portuguese (ABSAPT) 2022 at IberLEF 2022. We submitted the best
performing systems, achieving new state-of-the-art results on both subtasks. | cs.CL | Deep Learning Brasil at ABSAPT 2022: Portuguese Transformer Ensemble Approaches |
2023-11-08T22:42:31Z | http://arxiv.org/pdf/2311.05047v1 | http://arxiv.org/abs/2311.05047v1 | Eduardo Garcia, Juliana Gomes, Adalberto Barbosa Júnior, Cardeque Borges, Nádia da Silva | 2023-11-08T22:42:31Z | cs.CL, cs.AI | In this paper, we delineate the strategy employed by our team,
DeepLearningBrasil, which secured us the first place in the shared task
DepSign-LT-EDI@RANLP-2023, achieving a 47.0% Macro F1-Score and a notable 2.4%
advantage. The task was to classify social media texts into three distinct
levels of depression - "not depressed," "moderately depressed," and "severely
depressed." Leveraging the power of the RoBERTa and DeBERTa models, we further
pre-trained them on a collected Reddit dataset, specifically curated from
mental health-related Reddit's communities (Subreddits), leading to an enhanced
understanding of nuanced mental health discourse. To address lengthy textual
data, we used truncation techniques that retained the essence of the content by
focusing on its beginnings and endings. Our model was robust against unbalanced
data by incorporating sample weights into the loss. Cross-validation and
ensemble techniques were then employed to combine our k-fold trained models,
delivering an optimal solution. The accompanying code is made available for
transparency and further development. | cs.CL | DeepLearningBrasil@LT-EDI-2023: Exploring Deep Learning Techniques for Detecting Depression in Social Media Text |
2023-11-08T22:18:53Z | http://arxiv.org/pdf/2311.05043v1 | http://arxiv.org/abs/2311.05043v1 | Leonard Salewski, A. Sophia Koepke, Hendrik P. A. Lensch, Zeynep Akata | 2023-11-08T22:18:53Z | cs.CV, cs.AI, cs.CL | Converting a model's internals to text can yield human-understandable
insights about the model. Inspired by the recent success of training-free
approaches for image captioning, we propose ZS-A2T, a zero-shot framework that
translates the transformer attention of a given model into natural language
without requiring any training. We consider this in the context of Visual
Question Answering (VQA). ZS-A2T builds on a pre-trained large language model
(LLM), which receives a task prompt, question, and predicted answer, as inputs.
The LLM is guided to select tokens which describe the regions in the input
image that the VQA model attended to. Crucially, we determine this similarity
by exploiting the text-image matching capabilities of the underlying VQA model.
Our framework does not require any training and allows the drop-in replacement
of different guiding sources (e.g. attribution instead of attention maps), or
language models. We evaluate this novel task on textual explanation datasets
for VQA, giving state-of-the-art performances for the zero-shot setting on
GQA-REX and VQA-X. Our code is available at:
https://github.com/ExplainableML/ZS-A2T. | cs.CV | Zero-shot Translation of Attention Patterns in VQA Models to Natural Language |
2023-11-08T21:41:45Z | http://arxiv.org/pdf/2311.07588v1 | http://arxiv.org/abs/2311.07588v1 | Ruijie Wang, Zhiruo Zhang, Luca Rossetto, Florian Ruosch, Abraham Bernstein | 2023-11-08T21:41:45Z | cs.CL, cs.AI | In recent years, scholarly data has grown dramatically in terms of both scale
and complexity. It becomes increasingly challenging to retrieve information
from scholarly knowledge graphs that include large-scale heterogeneous
relationships, such as authorship, affiliation, and citation, between various
types of entities, e.g., scholars, papers, and organizations. As part of the
Scholarly QALD Challenge, this paper presents a question-answering (QA) system
called NLQxform, which provides an easy-to-use natural language interface to
facilitate accessing scholarly knowledge graphs. NLQxform allows users to
express their complex query intentions in natural language questions. A
transformer-based language model, i.e., BART, is employed to translate
questions into standard SPARQL queries, which can be evaluated to retrieve the
required information. According to the public leaderboard of the Scholarly QALD
Challenge at ISWC 2023 (Task 1: DBLP-QUAD - Knowledge Graph Question Answering
over DBLP), NLQxform achieved an F1 score of 0.85 and ranked first on the QA
task, demonstrating the competitiveness of the system. | cs.CL | NLQxform: A Language Model-based Question to SPARQL Transformer |
2023-11-08T21:13:38Z | http://arxiv.org/pdf/2311.05020v2 | http://arxiv.org/abs/2311.05020v2 | Naomi Saphra, Eve Fleisig, Kyunghyun Cho, Adam Lopez | 2024-03-26T02:23:27Z | cs.CL | Many NLP researchers are experiencing an existential crisis triggered by the
astonishing success of ChatGPT and other systems based on large language models
(LLMs). After such a disruptive change to our understanding of the field, what
is left to do? Taking a historical lens, we look for guidance from the first
era of LLMs, which began in 2005 with large $n$-gram models for machine
translation (MT). We identify durable lessons from the first era, and more
importantly, we identify evergreen problems where NLP researchers can continue
to make meaningful contributions in areas where LLMs are ascendant. We argue
that disparities in scale are transient and researchers can work to reduce
them; that data, rather than hardware, is still a bottleneck for many
applications; that meaningful realistic evaluation is still an open problem;
and that there is still room for speculative approaches. | cs.CL | First Tragedy, then Parse: History Repeats Itself in the New Era of Large Language Models |
2023-11-08T20:41:18Z | http://arxiv.org/pdf/2311.05014v1 | http://arxiv.org/abs/2311.05014v1 | Zhen Tan, Lu Cheng, Song Wang, Yuan Bo, Jundong Li, Huan Liu | 2023-11-08T20:41:18Z | cs.CL, cs.AI | Pretrained language models (PLMs) have made significant strides in various
natural language processing tasks. However, the lack of interpretability due to
their ``black-box'' nature poses challenges for responsible implementation.
Although previous studies have attempted to improve interpretability by using,
e.g., attention weights in self-attention layers, these weights often lack
clarity, readability, and intuitiveness. In this research, we propose a novel
approach to interpreting PLMs by employing high-level, meaningful concepts that
are easily understandable for humans. For example, we learn the concept of
``Food'' and investigate how it influences the prediction of a model's
sentiment towards a restaurant review. We introduce C$^3$M, which combines
human-annotated and machine-generated concepts to extract hidden neurons
designed to encapsulate semantically meaningful and task-specific concepts.
Through empirical evaluations on real-world datasets, we manifest that our
approach offers valuable insights to interpret PLM behavior, helps diagnose
model failures, and enhances model robustness amidst noisy concept labels. | cs.CL | Interpreting Pretrained Language Models via Concept Bottlenecks |
2023-11-08T19:07:10Z | http://arxiv.org/pdf/2311.07587v2 | http://arxiv.org/abs/2311.07587v2 | C. Daniel Freeman, Laura Culp, Aaron Parisi, Maxwell L Bileschi, Gamaleldin F Elsayed, Alex Rizkowsky, Isabelle Simpson, Alex Alemi, Azade Nova, Ben Adlam, Bernd Bohnet, Gaurav Mishra, Hanie Sedghi, Igor Mordatch, Izzeddin Gur, Jaehoon Lee, JD Co-Reyes, Jeffrey Pennington, Kelvin Xu, Kevin Swersky, Kshiteej Mahajan, Lechao Xiao, Rosanne Liu, Simon Kornblith, Noah Constant, Peter J. Liu, Roman Novak, Yundi Qian, Noah Fiedel, Jascha Sohl-Dickstein | 2023-11-15T19:49:02Z | cs.CL, cs.AI, cs.CY, cs.LG | We introduce and study the problem of adversarial arithmetic, which provides
a simple yet challenging testbed for language model alignment. This problem is
comprised of arithmetic questions posed in natural language, with an arbitrary
adversarial string inserted before the question is complete. Even in the simple
setting of 1-digit addition problems, it is easy to find adversarial prompts
that make all tested models (including PaLM2, GPT4, Claude2) misbehave, and
even to steer models to a particular wrong answer. We additionally provide a
simple algorithm for finding successful attacks by querying those same models,
which we name "prompt inversion rejection sampling" (PIRS). We finally show
that models can be partially hardened against these attacks via reinforcement
learning and via agentic constitutional loops. However, we were not able to
make a language model fully robust against adversarial arithmetic attacks. | cs.CL | Frontier Language Models are not Robust to Adversarial Arithmetic, or "What do I need to say so you agree 2+2=5? |
2023-11-08T19:01:13Z | http://arxiv.org/pdf/2311.04978v2 | http://arxiv.org/abs/2311.04978v2 | Junyi Li, Ninareh Mehrabi, Charith Peris, Palash Goyal, Kai-Wei Chang, Aram Galstyan, Richard Zemel, Rahul Gupta | 2024-04-02T18:29:52Z | cs.CL | Large language models (LLMs) are known to generate biased responses where the
opinions of certain groups and populations are underrepresented. Here, we
present a novel approach to achieve controllable generation of specific
viewpoints using LLMs, that can be leveraged to produce multiple perspectives
and to reflect the diverse opinions. Moving beyond the traditional reliance on
demographics like age, gender, or party affiliation, we introduce a data-driven
notion of persona grounded in collaborative filtering, which is defined as
either a single individual or a cohort of individuals manifesting similar views
across specific inquiries. As individuals in the same demographic group may
have different personas, our data-driven persona definition allows for a more
nuanced understanding of different (latent) social groups present in the
population. In addition to this, we also explore an efficient method to steer
LLMs toward the personas that we define. We show that our data-driven personas
significantly enhance model steerability, with improvements of between
$57\%-77\%$ over our best performing baselines. | cs.CL | On the steerability of large language models toward data-driven personas |
2023-11-08T18:59:54Z | http://arxiv.org/pdf/2311.04902v2 | http://arxiv.org/abs/2311.04902v2 | Rocktim Jyoti Das, Mingjie Sun, Liqun Ma, Zhiqiang Shen | 2024-04-08T22:42:49Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) with billions of parameters are prime targets
for network pruning, removing some model weights without hurting performance.
Prior approaches such as magnitude pruning, SparseGPT, and Wanda, either
concentrated solely on weights or integrated weights with activations for
sparsity. However, they overlooked the informative gradients derived from
pretrained LLMs. In this paper, we present a novel sparsity-centric pruning
method for pretrained LLMs, termed Gradient-based Language Model Pruner
(GBLM-Pruner). GBLM-Pruner leverages the first-order term of the Taylor
expansion, operating in a training-free manner by harnessing properly
normalized gradients from a few calibration samples to determine the pruning
metric, and substantially outperforms competitive counterparts like SparseGPT
and Wanda in multiple benchmarks. Intriguingly, by incorporating gradients,
unstructured pruning with our method tends to reveal some structural patterns,
which mirrors the geometric interdependence inherent in the LLMs' parameter
structure. Additionally, GBLM-Pruner functions without any subsequent
retraining or weight updates to maintain its simplicity as other counterparts.
Extensive evaluations on LLaMA-1 and LLaMA-2 across various benchmarks show
that GBLM-Pruner surpasses magnitude pruning, Wanda and SparseGPT by
significant margins. We further extend our approach on Vision Transformer. Our
code and models are available at https://github.com/VILA-Lab/GBLM-Pruner. | cs.CL | Beyond Size: How Gradients Shape Pruning Decisions in Large Language Models |
2023-11-08T18:58:43Z | http://arxiv.org/pdf/2311.04900v1 | http://arxiv.org/abs/2311.04900v1 | Michael Wilson, Jackson Petty, Robert Frank | 2023-11-08T18:58:43Z | cs.CL | Language models are typically evaluated on their success at predicting the
distribution of specific words in specific contexts. Yet linguistic knowledge
also encodes relationships between contexts, allowing inferences between word
distributions. We investigate the degree to which pre-trained Transformer-based
large language models (LLMs) represent such relationships, focusing on the
domain of argument structure. We find that LLMs perform well in generalizing
the distribution of a novel noun argument between related contexts that were
seen during pre-training (e.g., the active object and passive subject of the
verb spray), succeeding by making use of the semantically-organized structure
of the embedding space for word embeddings. However, LLMs fail at
generalizations between related contexts that have not been observed during
pre-training, but which instantiate more abstract, but well-attested structural
generalizations (e.g., between the active object and passive subject of an
arbitrary verb). Instead, in this case, LLMs show a bias to generalize based on
linear order. This finding points to a limitation with current models and
points to a reason for which their training is data-intensive.s reported here
are available at https://github.com/clay-lab/structural-alternations. | cs.CL | How Abstract Is Linguistic Generalization in Large Language Models? Experiments with Argument Structure |
2023-11-08T18:57:23Z | http://arxiv.org/pdf/2311.04954v1 | http://arxiv.org/abs/2311.04954v1 | Luca Beurer-Kellner, Mark Niklas Müller, Marc Fischer, Martin Vechev | 2023-11-08T18:57:23Z | cs.CL, cs.AI | Many recent prompting strategies for large language models (LLMs) query the
model multiple times sequentially -- first to produce intermediate results and
then the final answer. However, using these methods, both decoder and model are
unaware of potential follow-up prompts, leading to disconnected and undesirably
wordy intermediate responses. In this work, we address this issue by proposing
prompt sketching, a new prompting paradigm in which an LLM does not only
respond by completing a prompt, but by predicting values for multiple variables
in a template. This way, sketching grants users more control over the
generation process, e.g., by providing a reasoning framework via intermediate
instructions, leading to better overall results. The key idea enabling
sketching with existing, autoregressive models is to adapt the decoding
procedure to also score follow-up instructions during text generation, thus
optimizing overall template likelihood in inference. Our experiments show that
in a zero-shot setting, prompt sketching outperforms existing, sequential
prompting schemes such as direct asking or chain-of-thought on 7 out of 8 LLM
benchmarking tasks, including state tracking, arithmetic reasoning, and general
question answering. To facilitate future use, we release a number of generic,
yet effective sketches applicable to many tasks, and an open source library
called dclib, powering our sketch-aware decoders. | cs.CL | Prompt Sketching for Large Language Models |
2023-11-08T18:56:35Z | http://arxiv.org/pdf/2311.04897v1 | http://arxiv.org/abs/2311.04897v1 | Koyena Pal, Jiuding Sun, Andrew Yuan, Byron C. Wallace, David Bau | 2023-11-08T18:56:35Z | cs.CL, cs.LG | We conjecture that hidden state vectors corresponding to individual input
tokens encode information sufficient to accurately predict several tokens
ahead. More concretely, in this paper we ask: Given a hidden (internal)
representation of a single token at position $t$ in an input, can we reliably
anticipate the tokens that will appear at positions $\geq t + 2$? To test this,
we measure linear approximation and causal intervention methods in GPT-J-6B to
evaluate the degree to which individual hidden states in the network contain
signal rich enough to predict future hidden states and, ultimately, token
outputs. We find that, at some layers, we can approximate a model's output with
more than 48% accuracy with respect to its prediction of subsequent tokens
through a single hidden state. Finally we present a "Future Lens" visualization
that uses these methods to create a new view of transformer states. | cs.CL | Future Lens: Anticipating Subsequent Tokens from a Single Hidden State |
2023-11-08T18:52:17Z | http://arxiv.org/pdf/2311.04892v2 | http://arxiv.org/abs/2311.04892v2 | Shashank Gupta, Vaishnavi Shrivastava, Ameet Deshpande, Ashwin Kalyan, Peter Clark, Ashish Sabharwal, Tushar Khot | 2024-01-27T08:49:29Z | cs.CL | Recent works have showcased the ability of LLMs to embody diverse personas in
their responses, exemplified by prompts like 'You are Yoda. Explain the Theory
of Relativity.' While this ability allows personalization of LLMs and enables
human behavior simulation, its effect on LLMs' capabilities remains unclear. To
fill this gap, we present the first extensive study of the unintended
side-effects of persona assignment on the ability of LLMs to perform basic
reasoning tasks. Our study covers 24 reasoning datasets, 4 LLMs, and 19 diverse
personas (e.g. an Asian person) spanning 5 socio-demographic groups. Our
experiments unveil that LLMs harbor deep rooted bias against various
socio-demographics underneath a veneer of fairness. While they overtly reject
stereotypes when explicitly asked ('Are Black people less skilled at
mathematics?'), they manifest stereotypical and erroneous presumptions when
asked to answer questions while adopting a persona. These can be observed as
abstentions in responses, e.g., 'As a Black person, I can't answer this
question as it requires math knowledge', and generally result in a substantial
performance drop. Our experiments with ChatGPT-3.5 show that this bias is
ubiquitous - 80% of our personas demonstrate bias; it is significant - some
datasets show performance drops of 70%+; and can be especially harmful for
certain groups - some personas suffer statistically significant drops on 80%+
of the datasets. Overall, all 4 LLMs exhibit this bias to varying extents, with
GPT-4-Turbo showing the least but still a problematic amount of bias (evident
in 42% of the personas). Further analysis shows that these persona-induced
errors can be hard-to-discern and hard-to-avoid. Our findings serve as a
cautionary tale that the practice of assigning personas to LLMs - a trend on
the rise - can surface their deep-rooted biases and have unforeseeable and
detrimental side-effects. | cs.CL | Bias Runs Deep: Implicit Reasoning Biases in Persona-Assigned LLMs |
2023-11-08T18:46:32Z | http://arxiv.org/pdf/2311.04886v2 | http://arxiv.org/abs/2311.04886v2 | Tal Schuster, Adam D. Lelkes, Haitian Sun, Jai Gupta, Jonathan Berant, William W. Cohen, Donald Metzler | 2024-06-30T18:53:22Z | cs.CL, cs.AI, cs.LG | Recently proposed long-form question answering (QA) systems, supported by
large language models (LLMs), have shown promising capabilities. Yet,
attributing and verifying their generated abstractive answers can be difficult,
and automatically evaluating their accuracy remains an ongoing challenge.
In this work, we introduce a new QA task for answering multi-answer questions
by summarizing multiple diverse sources in a semi-extractive fashion.
Specifically, Semi-extractive Multi-source QA (SEMQA) requires models to output
a comprehensive answer, while mixing factual quoted spans -- copied verbatim
from given input sources -- and non-factual free-text connectors that glue
these spans together into a single cohesive passage. This setting bridges the
gap between the outputs of well-grounded but constrained extractive QA systems
and more fluent but harder to attribute fully abstractive answers.
Particularly, it enables a new mode for language models that leverages their
advanced language generation capabilities, while also producing fine in-line
attributions by-design that are easy to verify, interpret, and evaluate.
To study this task, we create the first dataset of this kind, QuoteSum, with
human-written semi-extractive answers to natural and generated questions, and
define text-based evaluation metrics. Experimenting with several LLMs in
various settings, we find this task to be surprisingly challenging,
demonstrating the importance of QuoteSum for developing and studying such
consolidation capabilities. | cs.CL | SEMQA: Semi-Extractive Multi-Source Question Answering |
2023-11-08T18:44:47Z | http://arxiv.org/pdf/2311.04885v1 | http://arxiv.org/abs/2311.04885v1 | Tibor L. R. Krols, Marie Mortensen, Ninell Oldenburg | 2023-11-08T18:44:47Z | cs.CL | Social media has become a very popular source of information. With this
popularity comes an interest in systems that can classify the information
produced. This study tries to create such a system detecting irony in Twitter
users. Recent work emphasize the importance of lexical features, sentiment
features and the contrast herein along with TF-IDF and topic models. Based on a
thorough feature selection process, the resulting model contains specific
sub-features from these areas. Our model reaches an F1-score of 0.84, which is
above the baseline. We find that lexical features, especially TF-IDF,
contribute the most to our models while sentiment and topic modeling features
contribute less to overall performance. Lastly, we highlight multiple
interesting and important paths for further exploration. | cs.CL | Profiling Irony & Stereotype: Exploring Sentiment, Topic, and Lexical Features |
2023-11-08T18:33:06Z | http://arxiv.org/pdf/2311.04879v2 | http://arxiv.org/abs/2311.04879v2 | Jianxin Yang | 2023-11-09T05:21:28Z | cs.CL, cs.AI | We present LongQLoRA, an efficient and effective method to extend context
length of large language models with less training resources. LongQLoRA
combines the advantages of Position Interpolation, QLoRA and Shift Short
Attention of LongLoRA. With a single 32GB V100 GPU, LongQLoRA can extend the
context length of LLaMA2 7B and 13B from 4096 to 8192 and even to 12k within
1000 finetuning steps. LongQLoRA achieves competitive perplexity performance on
PG19 and Proof-pile datasets, our model outperforms LongLoRA and is very close
to MPT-7B-8K within the evaluation context length of 8192. We collect and build
39k long instruction data to extend context length of Vicuna-13B from 4096 to
8192 and achieve good performance both in long and short context generation
task. We also do some ablation experiments to study the effect of LoRA rank,
finetuning steps and attention patterns in inference.The model weights,
training data and code are avaliable at
https://github.com/yangjianxin1/LongQLoRA. | cs.CL | LongQLoRA: Efficient and Effective Method to Extend Context Length of Large Language Models |
2023-11-08T17:59:15Z | http://arxiv.org/pdf/2311.05772v2 | http://arxiv.org/abs/2311.05772v2 | Archiki Prasad, Alexander Koller, Mareike Hartmann, Peter Clark, Ashish Sabharwal, Mohit Bansal, Tushar Khot | 2024-04-08T20:42:17Z | cs.AI, cs.CL, cs.LG | Large Language Models (LLMs) are increasingly being used for interactive
decision-making tasks requiring planning and adapting to the environment.
Recent works employ LLMs-as-agents in broadly two ways: iteratively determining
the next action (iterative executors) or generating plans and executing
sub-tasks using LLMs (plan-and-execute). However, these methods struggle with
task complexity, as the inability to execute any sub-task may lead to task
failure. To address these shortcomings, we introduce As-Needed Decomposition
and Planning for complex Tasks (ADaPT), an approach that explicitly plans and
decomposes complex sub-tasks as-needed, i.e., when the LLM is unable to execute
them. ADaPT recursively decomposes sub-tasks to adapt to both task complexity
and LLM capability. Our results demonstrate that ADaPT substantially
outperforms established strong baselines, achieving success rates up to 28.3%
higher in ALFWorld, 27% in WebShop, and 33% in TextCraft -- a novel
compositional dataset that we introduce. Through extensive analysis, we
illustrate the importance of multilevel decomposition and establish that ADaPT
dynamically adjusts to the capabilities of the executor LLM as well as to task
complexity. | cs.AI | ADaPT: As-Needed Decomposition and Planning with Language Models |
2023-11-08T17:35:20Z | http://arxiv.org/pdf/2311.04850v2 | http://arxiv.org/abs/2311.04850v2 | Shuo Yang, Wei-Lin Chiang, Lianmin Zheng, Joseph E. Gonzalez, Ion Stoica | 2023-11-11T05:11:18Z | cs.CL, cs.AI | Large language models are increasingly trained on all the data ever produced
by humans. Many have raised concerns about the trustworthiness of public
benchmarks due to potential contamination in pre-training or fine-tuning
datasets. While most data decontamination efforts apply string matching (e.g.,
n-gram overlap) to remove benchmark data, we show that these methods are
insufficient, and simple variations of test data (e.g., paraphrasing,
translation) can easily bypass these decontamination measures. Furthermore, we
demonstrate that if such variation of test data is not eliminated, a 13B model
can easily overfit a test benchmark and achieve drastically high performance,
on par with GPT-4. We validate such observations in widely used benchmarks such
as MMLU, GSK8k, and HumanEval. To address this growing risk, we propose a
stronger LLM-based decontamination method and apply it to widely used
pre-training and fine-tuning datasets, revealing significant previously unknown
test overlap. For example, in pre-training sets such as RedPajama-Data-1T and
StarCoder-Data, we identified that 8-18\% of the HumanEval benchmark overlaps.
Interestingly, we also find such contamination in synthetic dataset generated
by GPT-3.5/4, suggesting a potential risk of unintentional contamination. We
urge the community to adopt stronger decontamination approaches when using
public benchmarks. Moreover, we call for the community to actively develop
fresh one-time exams to evaluate models accurately. Our decontamination tool is
publicly available at https://github.com/lm-sys/llm-decontaminator. | cs.CL | Rethinking Benchmark and Contamination for Language Models with Rephrased Samples |
2023-11-08T16:50:05Z | http://arxiv.org/pdf/2311.04823v1 | http://arxiv.org/abs/2311.04823v1 | Zhen Qin, Songlin Yang, Yiran Zhong | 2023-11-08T16:50:05Z | cs.CL, cs.LG | Transformers have surpassed RNNs in popularity due to their superior
abilities in parallel training and long-term dependency modeling. Recently,
there has been a renewed interest in using linear RNNs for efficient sequence
modeling. These linear RNNs often employ gating mechanisms in the output of the
linear recurrence layer while ignoring the significance of using forget gates
within the recurrence. In this paper, we propose a gated linear RNN model
dubbed Hierarchically Gated Recurrent Neural Network (HGRN), which includes
forget gates that are lower bounded by a learnable value. The lower bound
increases monotonically when moving up layers. This allows the upper layers to
model long-term dependencies and the lower layers to model more local,
short-term dependencies. Experiments on language modeling, image
classification, and long-range arena benchmarks showcase the efficiency and
effectiveness of our proposed model. The source code is available at
https://github.com/OpenNLPLab/HGRN. | cs.CL | Hierarchically Gated Recurrent Neural Network for Sequence Modeling |
2023-11-08T16:41:37Z | http://arxiv.org/pdf/2311.04816v1 | http://arxiv.org/abs/2311.04816v1 | Zheng Chu, Zekun Wang, Jiafeng Liang, Ming Liu, Bing Qin | 2023-11-08T16:41:37Z | cs.CL, cs.AI | The facts and time in the document are intricately intertwined, making
temporal reasoning over documents challenging. Previous work models time
implicitly, making it difficult to handle such complex relationships. To
address this issue, we propose MTGER, a novel Multi-view Temporal Graph
Enhanced Temporal Reasoning framework for temporal reasoning over time-involved
documents. Concretely, MTGER explicitly models the temporal relationships among
facts by multi-view temporal graphs. On the one hand, the heterogeneous
temporal graphs explicitly model the temporal and discourse relationships among
facts; on the other hand, the multi-view mechanism captures both time-focused
and fact-focused information, allowing the two views to complement each other
through adaptive fusion. To further improve the implicit reasoning capability
of the model, we design a self-supervised time-comparing objective. Extensive
experimental results demonstrate the effectiveness of our method on the TimeQA
and SituatedQA datasets. Furthermore, MTGER gives more consistent answers under
question perturbations. | cs.CL | MTGER: Multi-view Temporal Graph Enhanced Temporal Reasoning over Time-Involved Document |
2023-11-08T16:18:32Z | http://arxiv.org/pdf/2311.04799v1 | http://arxiv.org/abs/2311.04799v1 | Martin Kuo, Jianyi Zhang, Yiran Chen | 2023-11-08T16:18:32Z | cs.CL, cs.AI | Building on the cost-efficient pretraining advancements brought about by
Crammed BERT, we enhance its performance and interpretability further by
introducing a novel pretrained model Dependency Agreement Crammed BERT
(DACBERT) and its two-stage pretraining framework - Dependency Agreement
Pretraining. This framework, grounded by linguistic theories, seamlessly weaves
syntax and semantic information into the pretraining process. The first stage
employs four dedicated submodels to capture representative dependency
agreements at the chunk level, effectively converting these agreements into
embeddings. The second stage uses these refined embeddings, in tandem with
conventional BERT embeddings, to guide the pretraining of the rest of the
model. Evaluated on the GLUE benchmark, our DACBERT demonstrates notable
improvement across various tasks, surpassing Crammed BERT by 3.13% in the RTE
task and by 2.26% in the MRPC task. Furthermore, our method boosts the average
GLUE score by 0.83%, underscoring its significant potential. The pretraining
process can be efficiently executed on a single GPU within a 24-hour cycle,
necessitating no supplementary computational resources or extending the
pretraining duration compared with the Crammed BERT. Extensive studies further
illuminate our approach's instrumental role in bolstering the interpretability
of pretrained language models for natural language understanding tasks. | cs.CL | DACBERT: Leveraging Dependency Agreement for Cost-Efficient Bert Pretraining |
2023-11-08T16:10:28Z | http://arxiv.org/pdf/2311.04789v1 | http://arxiv.org/abs/2311.04789v1 | Md Azim Khan | 2023-11-08T16:10:28Z | cs.LG, cs.CL, cs.CY | Online conversations can be toxic and subjected to threats, abuse, or
harassment. To identify toxic text comments, several deep learning and machine
learning models have been proposed throughout the years. However, recent
studies demonstrate that because of the imbalances in the training data, some
models are more likely to show unintended biases including gender bias and
identity bias. In this research, our aim is to detect toxic comment and reduce
the unintended bias concerning identity features such as race, gender, sex,
religion by fine-tuning an attention based model called BERT(Bidirectional
Encoder Representation from Transformers). We apply weighted loss to address
the issue of unbalanced data and compare the performance of a fine-tuned BERT
model with a traditional Logistic Regression model in terms of classification
and bias minimization. The Logistic Regression model with the TFIDF vectorizer
achieve 57.1% accuracy, and fine-tuned BERT model's accuracy is 89%. Code is
available at
https://github.com/zim10/Determine_Toxic_comment_and_identity_bias.git | cs.LG | Determination of toxic comments and unintended model bias minimization using Deep learning approach |
2023-11-08T15:11:57Z | http://arxiv.org/pdf/2311.04742v2 | http://arxiv.org/abs/2311.04742v2 | Antonios Georgiou, Tankut Can, Mikhail Katkov, Misha Tsodyks | 2023-11-28T05:25:45Z | cs.CL, q-bio.NC | One of the most impressive achievements of the AI revolution is the
development of large language models that can generate meaningful text and
respond to instructions in plain English with no additional training necessary.
Here we show that language models can be used as a scientific instrument for
studying human memory for meaningful material. We developed a pipeline for
designing large scale memory experiments and analyzing the obtained results. We
performed online memory experiments with a large number of participants and
collected recognition and recall data for narratives of different lengths. We
found that both recall and recognition performance scale linearly with
narrative length. Furthermore, in order to investigate the role of narrative
comprehension in memory, we repeated these experiments using scrambled versions
of the presented stories. We found that even though recall performance declined
significantly, recognition remained largely unaffected. Interestingly, recalls
in this condition seem to follow the original narrative order rather than the
scrambled presentation, pointing to a contextual reconstruction of the story in
memory. | cs.CL | Using large language models to study human memory for meaningful narratives |
2023-11-08T14:05:00Z | http://arxiv.org/pdf/2311.04694v3 | http://arxiv.org/abs/2311.04694v3 | Lukas Gienapp, Harrisen Scells, Niklas Deckers, Janek Bevendorff, Shuai Wang, Johannes Kiesel, Shahbaz Syed, Maik Fröbe, Guido Zuccon, Benno Stein, Matthias Hagen, Martin Potthast | 2024-05-22T10:33:56Z | cs.IR, cs.CL | Recent advances in large language models have enabled the development of
viable generative retrieval systems. Instead of a traditional document ranking,
generative retrieval systems often directly return a grounded generated text as
a response to a query. Quantifying the utility of the textual responses is
essential for appropriately evaluating such generative ad hoc retrieval. Yet,
the established evaluation methodology for ranking-based ad hoc retrieval is
not suited for the reliable and reproducible evaluation of generated responses.
To lay a foundation for developing new evaluation methods for generative
retrieval systems, we survey the relevant literature from the fields of
information retrieval and natural language processing, identify search tasks
and system architectures in generative retrieval, develop a new user model, and
study its operationalization. | cs.IR | Evaluating Generative Ad Hoc Information Retrieval |
2023-11-08T13:13:23Z | http://arxiv.org/pdf/2311.04666v4 | http://arxiv.org/abs/2311.04666v4 | Khushi Bhardwaj, Raj Sanjay Shah, Sashank Varma | 2024-01-10T05:36:05Z | cs.CL, cs.AI | Pre-trained Large Language Models (LLMs) have shown success in a diverse set
of language inference and understanding tasks. The pre-training stage of LLMs
looks at a large corpus of raw textual data. The BabyLM shared task compares
LLM pre-training to human language acquisition, where the number of tokens seen
by 13-year-old kids is magnitudes smaller than the number of tokens seen by
LLMs. In this work, we pre-train and evaluate LLMs on their ability to learn
contextual word representations using roughly the same number of tokens as seen
by children. We provide a strong set of baselines; with different
architectures, evaluation of changes in performance across epochs, and reported
pre-training metrics for the strict small and strict tracks of the task. We
also try to loosely replicate the RoBERTa baseline given by the task organizers
to observe the training robustness to hyperparameter selection and
replicability. We provide the submission details to the strict and strict-small
tracks in this report. | cs.CL | Pre-training LLMs using human-like development data corpus |
2023-11-08T13:11:48Z | http://arxiv.org/pdf/2311.04664v2 | http://arxiv.org/abs/2311.04664v2 | Subba Reddy Oota, Emin Çelik, Fatma Deniz, Mariya Toneva | 2024-06-16T23:52:21Z | cs.CL, cs.LG, eess.AS, q-bio.NC | Despite known differences between reading and listening in the brain, recent
work has shown that text-based language models predict both text-evoked and
speech-evoked brain activity to an impressive degree. This poses the question
of what types of information language models truly predict in the brain. We
investigate this question via a direct approach, in which we systematically
remove specific low-level stimulus features (textual, speech, and visual) from
language model representations to assess their impact on alignment with fMRI
brain recordings during reading and listening. Comparing these findings with
speech-based language models reveals starkly different effects of low-level
features on brain alignment. While text-based models show reduced alignment in
early sensory regions post-removal, they retain significant predictive power in
late language regions. In contrast, speech-based models maintain strong
alignment in early auditory regions even after feature removal but lose all
predictive power in late language regions. These results suggest that
speech-based models provide insights into additional information processed by
early auditory regions, but caution is needed when using them to model
processing in late language regions. We make our code publicly available.
[https://github.com/subbareddy248/speech-llm-brain] | cs.CL | Speech language models lack important brain-relevant semantics |
2023-11-08T13:03:06Z | http://arxiv.org/pdf/2311.04661v3 | http://arxiv.org/abs/2311.04661v3 | Chenmien Tan, Ge Zhang, Jie Fu | 2024-01-25T03:50:57Z | cs.CL, cs.LG | While large language models (LLMs) have enabled learning knowledge from the
pre-training corpora, the acquired knowledge may be fundamentally incorrect or
outdated over time, which necessitates rectifying the knowledge of the language
model (LM) after the training. A promising approach involves employing a
hyper-network to generate parameter shift, whereas existing hyper-networks
suffer from inferior scalability in synchronous editing operation amount. To
mitigate the problem, we propose the MAssive Language Model Editing Network
(MALMEN), which formulates the parameter shift aggregation as the least square
problem, subsequently updating the LM parameters using the normal equation. To
accommodate editing multiple facts simultaneously with limited memory budgets,
we separate the computation on the hyper-network and LM, enabling arbitrary
batch size on both neural networks. Our method is evaluated by editing up to
thousands of facts on LMs with different architectures, i.e., BERT-base, GPT-2,
T5-XL (2.8B), and GPT-J (6B), across various knowledge-intensive NLP tasks,
i.e., closed book fact-checking and question answering. Remarkably, MALMEN is
capable of editing hundreds of times more facts than strong baselines with the
identical hyper-network architecture and outperforms editor specifically
designed for GPT. Our code is available at
https://github.com/ChenmienTan/malmen. | cs.CL | Massive Editing for Large Language Models via Meta Learning |
2023-11-08T11:51:47Z | http://arxiv.org/pdf/2311.04948v1 | http://arxiv.org/abs/2311.04948v1 | David Novoa-Paradela, Oscar Fontenla-Romero, Bertha Guijarro-Berdiñas | 2023-11-08T11:51:47Z | cs.CL, cs.AI, cs.LG | This paper presents a pipeline to detect and explain anomalous reviews in
online platforms. The pipeline is made up of three modules and allows the
detection of reviews that do not generate value for users due to either
worthless or malicious composition. The classifications are accompanied by a
normality score and an explanation that justifies the decision made. The
pipeline's ability to solve the anomaly detection task was evaluated using
different datasets created from a large Amazon database. Additionally, a study
comparing three explainability techniques involving 241 participants was
conducted to assess the explainability module. The study aimed to measure the
impact of explanations on the respondents' ability to reproduce the
classification model and their perceived usefulness. This work can be useful to
automate tasks in review online platforms, such as those for electronic
commerce, and offers inspiration for addressing similar problems in the field
of anomaly detection in textual data. We also consider it interesting to have
carried out a human evaluation of the capacity of different explainability
techniques in a real and infrequent scenario such as the detection of anomalous
reviews, as well as to reflect on whether it is possible to explain tasks as
humanly subjective as this one. | cs.CL | Explained anomaly detection in text reviews: Can subjective scenarios be correctly evaluated? |
2023-11-08T10:34:16Z | http://arxiv.org/pdf/2311.04589v3 | http://arxiv.org/abs/2311.04589v3 | Zhen Yang, Yingxue Zhang, Fandong Meng, Jie Zhou | 2024-01-04T07:31:07Z | cs.CL, cs.AI | Despite Multi-modal Large Language Models (MM-LLMs) have made exciting
strides recently, they are still struggling to efficiently model the
interactions among multi-modal inputs and the generation in non-textual
modalities. In this work, we propose TEAL (Tokenize and Embed ALl)}, an
approach to treat the input from any modality as a token sequence and learn a
joint embedding space for all modalities. Specifically, for the input from any
modality, TEAL first discretizes it into a token sequence with the
off-the-shelf tokenizer and embeds the token sequence into a joint embedding
space with a learnable embedding matrix. MM-LLMs just need to predict the
multi-modal tokens autoregressively as the textual LLMs do. Finally, the
corresponding de-tokenizer is applied to generate the output in each modality
based on the predicted token sequence. With the joint embedding space, TEAL
enables the frozen LLMs to perform both understanding and generation tasks
involving non-textual modalities, such as image and audio. Thus, the textual
LLM can just work as an interface and maintain its high performance in textual
understanding and generation. Experiments show that TEAL achieves substantial
improvements in multi-modal understanding, and implements a simple scheme for
multi-modal generations. | cs.CL | TEAL: Tokenize and Embed ALL for Multi-modal Large Language Models |
2023-11-08T09:52:58Z | http://arxiv.org/pdf/2311.04563v1 | http://arxiv.org/abs/2311.04563v1 | Urban Knupleš, Diego Frassinelli, Sabine Schulte im Walde | 2023-11-08T09:52:58Z | cs.CL | Humans tend to strongly agree on ratings on a scale for extreme cases (e.g.,
a CAT is judged as very concrete), but judgements on mid-scale words exhibit
more disagreement. Yet, collected rating norms are heavily exploited across
disciplines. Our study focuses on concreteness ratings and (i) implements
correlations and supervised classification to identify salient multi-modal
characteristics of mid-scale words, and (ii) applies a hard clustering to
identify patterns of systematic disagreement across raters. Our results suggest
to either fine-tune or filter mid-scale target words before utilising them. | cs.CL | Investigating the Nature of Disagreements on Mid-Scale Ratings: A Case Study on the Abstractness-Concreteness Continuum |
2023-11-08T09:37:09Z | http://arxiv.org/pdf/2311.04554v1 | http://arxiv.org/abs/2311.04554v1 | Vatsal Raina, Adian Liusie, Mark Gales | 2023-11-08T09:37:09Z | cs.CL | Multiple-choice tests are a common approach for assessing candidates'
comprehension skills. Standard multiple-choice reading comprehension exams
require candidates to select the correct answer option from a discrete set
based on a question in relation to a contextual passage. For appropriate
assessment, the distractor answer options must by definition be incorrect but
plausible and diverse. However, generating good quality distractors satisfying
these criteria is a challenging task for content creators. We propose automated
assessment metrics for the quality of distractors in multiple-choice reading
comprehension tests. Specifically, we define quality in terms of the
incorrectness, plausibility and diversity of the distractor options. We assess
incorrectness using the classification ability of a binary multiple-choice
reading comprehension system. Plausibility is assessed by considering the
distractor confidence - the probability mass associated with the distractor
options for a standard multi-class multiple-choice reading comprehension
system. Diversity is assessed by pairwise comparison of an embedding-based
equivalence metric between the distractors of a question. To further validate
the plausibility metric we compare against candidate distributions over
multiple-choice questions and agreement with a ChatGPT model's interpretation
of distractor plausibility and diversity. | cs.CL | Assessing Distractors in Multiple-Choice Tests |
2023-11-08T09:26:27Z | http://arxiv.org/pdf/2311.04547v1 | http://arxiv.org/abs/2311.04547v1 | Julius Steuer, Marius Mosbach, Dietrich Klakow | 2023-11-08T09:26:27Z | cs.CL | Research on the cognitive plausibility of language models (LMs) has so far
mostly concentrated on modelling psycholinguistic response variables such as
reading times, gaze durations and N400/P600 EEG signals, while mostly leaving
out the dimension of what Mahowald et al. (2023) described as formal and
functional linguistic competence, and developmental plausibility. We address
this gap by training a series of GPT-like language models of different sizes on
the strict version of the BabyLM pretraining corpus, evaluating on the
challenge tasks (BLiMP, GLUE, MSGS) and an additional reading time prediction
task. We find a positive correlation between LM size and performance on all
three challenge tasks, with different preferences for model width and depth in
each of the tasks. In contrast, a negative correlation was found between LM
size and reading time fit of linear mixed-effects models using LM surprisal as
a predictor, with the second-smallest LM achieving the largest log-likelihood
reduction over a baseline model without surprisal. This suggests that modelling
processing effort and linguistic competence may require an approach different
from training GPT-like LMs on a developmentally plausible corpus. | cs.CL | Large GPT-like Models are Bad Babies: A Closer Look at the Relationship between Linguistic Competence and Psycholinguistic Measures |
2023-11-08T08:47:49Z | http://arxiv.org/pdf/2311.04535v1 | http://arxiv.org/abs/2311.04535v1 | Tiasa Singha Roy, Priyam Basu | 2023-11-08T08:47:49Z | cs.CL, cs.AI, cs.LG | Research on data generation and augmentation has been focused majorly on
enhancing generation models, leaving a notable gap in the exploration and
refinement of methods for evaluating synthetic data. There are several text
similarity metrics within the context of generated data filtering which can
impact the performance of specific Natural Language Understanding (NLU) tasks,
specifically focusing on intent and sentiment classification. In this study, we
propose RankAug, a text-ranking approach that detects and filters out the top
augmented texts in terms of being most similar in meaning with lexical and
syntactical diversity. Through experiments conducted on multiple datasets, we
demonstrate that the judicious selection of filtering techniques can yield a
substantial improvement of up to 35% in classification accuracy for
under-represented classes. | cs.CL | RankAug: Augmented data ranking for text classification |
2023-11-08T08:45:14Z | http://arxiv.org/pdf/2311.04534v2 | http://arxiv.org/abs/2311.04534v2 | Qian Chen, Wen Wang, Qinglin Zhang, Siqi Zheng, Shiliang Zhang, Chong Deng, Yukun Ma, Hai Yu, Jiaqing Liu, Chong Zhang | 2024-02-05T02:42:57Z | cs.CL, cs.SD, eess.AS | Recently, unified speech-text models, such as SpeechGPT, VioLA, and
AudioPaLM, have achieved remarkable performance on various speech tasks. These
models discretize speech signals into tokens (speech discretization) and use a
shared vocabulary for both text and speech tokens. Then they train a single
decoder-only Transformer on a mixture of speech tasks. However, these models
rely on the Loss Masking strategy for the ASR task, which ignores the
dependency among speech tokens. In this paper, we propose to model speech
tokens in an autoregressive way, similar to text. We find that applying the
conventional cross-entropy loss on input speech tokens does not consistently
improve the ASR performance over the Loss Masking approach. To address this
issue, we propose a novel approach denoted Smoothed Label Distillation (SLD),
which applies a KL divergence loss with smoothed labels on speech tokens. Our
experiments show that SLD effectively models speech tokens and outperforms Loss
Masking for decoder-only Transformers in ASR tasks with different speech
discretization methods. The source code can be found here:
https://github.com/alibaba-damo-academy/SpokenNLP/tree/main/sld | cs.CL | Loss Masking Is Not Needed in Decoder-only Transformer for Discrete-token-based ASR |
2023-11-08T07:46:25Z | http://arxiv.org/pdf/2311.04507v3 | http://arxiv.org/abs/2311.04507v3 | Cam-Van Thi Nguyen, Anh-Tuan Mai, The-Son Le, Hai-Dang Kieu, Duc-Trong Le | 2024-01-30T08:01:42Z | cs.CL, cs.MM | Emotion recognition is a crucial task for human conversation understanding.
It becomes more challenging with the notion of multimodal data, e.g., language,
voice, and facial expressions. As a typical solution, the global- and the local
context information are exploited to predict the emotional label for every
single sentence, i.e., utterance, in the dialogue. Specifically, the global
representation could be captured via modeling of cross-modal interactions at
the conversation level. The local one is often inferred using the temporal
information of speakers or emotional shifts, which neglects vital factors at
the utterance level. Additionally, most existing approaches take fused features
of multiple modalities in an unified input without leveraging modality-specific
representations. Motivating from these problems, we propose the Relational
Temporal Graph Neural Network with Auxiliary Cross-Modality Interaction
(CORECT), an novel neural network framework that effectively captures
conversation-level cross-modality interactions and utterance-level temporal
dependencies with the modality-specific manner for conversation understanding.
Extensive experiments demonstrate the effectiveness of CORECT via its
state-of-the-art results on the IEMOCAP and CMU-MOSEI datasets for the
multimodal ERC task. | cs.CL | Conversation Understanding using Relational Temporal Graph Neural Networks with Auxiliary Cross-Modality Interaction |
2023-11-08T07:15:05Z | http://arxiv.org/pdf/2311.04498v4 | http://arxiv.org/abs/2311.04498v4 | Ao Zhang, Yuan Yao, Wei Ji, Zhiyuan Liu, Tat-Seng Chua | 2023-12-18T12:15:26Z | cs.CV, cs.AI, cs.CL | The development of large language models (LLMs) has greatly advanced the
field of multimodal understanding, leading to the emergence of large multimodal
models (LMMs). In order to enhance the level of visual comprehension, recent
studies have equipped LMMs with region-level understanding capabilities by
representing object bounding box coordinates as a series of text sequences
(pix2seq). In this paper, we introduce a novel paradigm for object location
modeling called pix2emb method, where we ask the LMM to output the location
embeddings and then decode them with different decoders. This paradigm allows
us to use different location formats (such as bounding boxes and masks) in
multimodal conversations. Leveraging the proposed pix2emb method, we train an
LMM named NExT-Chat and demonstrate its capability of handling multiple tasks
like visual grounding, region captioning, and grounded reasoning. Comprehensive
experiments show the effectiveness of our NExT-Chat on various tasks, e.g.,
NExT-Chat (87.7) vs. Shikra (86.9) on POPE-Random, NExT-Chat (68.9) vs. LISA
(67.9) on referring expression segmentation task, and NExT-Chat (79.6) vs.
Kosmos-2 (62.3) on region caption task. The code and model are released at
https://github.com/NExT-ChatV/NExT-Chat. | cs.CV | NExT-Chat: An LMM for Chat, Detection and Segmentation |
2023-11-08T06:54:34Z | http://arxiv.org/pdf/2311.04495v1 | http://arxiv.org/abs/2311.04495v1 | Zhengyuan Liu, Hai Leong Chieu, Nancy F. Chen | 2023-11-08T06:54:34Z | cs.CL | Data collection from manual labeling provides domain-specific and
task-aligned supervision for data-driven approaches, and a critical mass of
well-annotated resources is required to achieve reasonable performance in
natural language processing tasks. However, manual annotations are often
challenging to scale up in terms of time and budget, especially when domain
knowledge, capturing subtle semantic features, and reasoning steps are needed.
In this paper, we investigate the efficacy of leveraging large language models
on automated labeling for computational stance detection. We empirically
observe that while large language models show strong potential as an
alternative to human annotators, their sensitivity to task-specific
instructions and their intrinsic biases pose intriguing yet unique challenges
in machine annotation. We introduce a multi-label and multi-target sampling
strategy to optimize the annotation quality. Experimental results on the
benchmark stance detection corpora show that our method can significantly
improve performance and learning efficacy. | cs.CL | Multi-label and Multi-target Sampling of Machine Annotation for Computational Stance Detection |
2023-11-08T06:20:32Z | http://arxiv.org/pdf/2311.04480v1 | http://arxiv.org/abs/2311.04480v1 | Cheng-Yu Chuang, Pooyan Fazli | 2023-11-08T06:20:32Z | cs.CV, cs.CL, cs.LG | Video description entails automatically generating coherent natural language
sentences that narrate the content of a given video. We introduce CLearViD, a
transformer-based model for video description generation that leverages
curriculum learning to accomplish this task. In particular, we investigate two
curriculum strategies: (1) progressively exposing the model to more challenging
samples by gradually applying a Gaussian noise to the video data, and (2)
gradually reducing the capacity of the network through dropout during the
training process. These methods enable the model to learn more robust and
generalizable features. Moreover, CLearViD leverages the Mish activation
function, which provides non-linearity and non-monotonicity and helps alleviate
the issue of vanishing gradients. Our extensive experiments and ablation
studies demonstrate the effectiveness of the proposed model. The results on two
datasets, namely ActivityNet Captions and YouCook2, show that CLearViD
significantly outperforms existing state-of-the-art models in terms of both
accuracy and diversity metrics. | cs.CV | CLearViD: Curriculum Learning for Video Description |
2023-11-08T06:16:04Z | http://arxiv.org/pdf/2311.04479v1 | http://arxiv.org/abs/2311.04479v1 | Wenbo Zhu, Tiechuan Hu | 2023-11-08T06:16:04Z | cs.CL, cs.IR, cs.LG, cs.SI | In this paper, we look at a database of tweets sorted by various keywords
that could indicate the users sentiment towards covid vaccines. With social
media becoming such a prevalent source of opinion, sorting and ranking tweets
that hold important information such as opinions on covid vaccines is of utmost
importance. Two different ranking scales were used, and ranking a tweet in this
way could represent the difference between an opinion being lost and an opinion
being featured on the site, which affects the decisions and behavior of people,
and why researchers were interested in it. Using natural language processing
techniques, our aim is to determine and categorize opinions about covid
vaccines with the highest accuracy possible. | cs.CL | Twitter Sentiment Analysis of Covid Vacciness |
2023-11-08T05:37:49Z | http://arxiv.org/pdf/2311.04467v1 | http://arxiv.org/abs/2311.04467v1 | Xusheng Zhao, Hao Peng, Qiong Dai, Xu Bai, Huailiang Peng, Yanbing Liu, Qinglang Guo, Philip S. Yu | 2023-11-08T05:37:49Z | cs.CL, cs.AI | Aspect-based sentiment analysis (ABSA) is dedicated to forecasting the
sentiment polarity of aspect terms within sentences. Employing graph neural
networks to capture structural patterns from syntactic dependency parsing has
been confirmed as an effective approach for boosting ABSA. In most works, the
topology of dependency trees or dependency-based attention coefficients is
often loosely regarded as edges between aspects and opinions, which can result
in insufficient and ambiguous syntactic utilization. To address these problems,
we propose a new reinforced dependency graph convolutional network (RDGCN) that
improves the importance calculation of dependencies in both distance and type
views. Initially, we propose an importance calculation criterion for the
minimum distances over dependency trees. Under the criterion, we design a
distance-importance function that leverages reinforcement learning for weight
distribution search and dissimilarity control. Since dependency types often do
not have explicit syntax like tree distances, we use global attention and mask
mechanisms to design type-importance functions. Finally, we merge these weights
and implement feature aggregation and classification. Comprehensive experiments
on three popular datasets demonstrate the effectiveness of the criterion and
importance functions. RDGCN outperforms state-of-the-art GNN-based baselines in
all validations. | cs.CL | RDGCN: Reinforced Dependency Graph Convolutional Network for Aspect-based Sentiment Analysis |
2023-11-08T04:58:29Z | http://arxiv.org/pdf/2311.04459v1 | http://arxiv.org/abs/2311.04459v1 | Yichen Wang, Kevin Yang, Xiaoming Liu, Dan Klein | 2023-11-08T04:58:29Z | cs.CL, cs.AI | Existing LLM-based systems for writing long-form stories or story outlines
frequently suffer from unnatural pacing, whether glossing over important events
or over-elaborating on insignificant details, resulting in a jarring experience
for the reader. We propose a CONCrete Outline ConTrol (CONCOCT) system to
improve pacing when automatically generating story outlines. We first train a
concreteness evaluator to judge which of two events is more concrete
(low-level-detailed). This evaluator can then be used to control pacing in
hierarchical outline generation; in this work, we explore a vaguest-first
expansion procedure that aims for uniform pacing. We further use the evaluator
to filter new outline items based on predicted concreteness. Compared to a
baseline hierarchical outline generator, humans judge CONCOCT's pacing to be
more consistent over 57% of the time across multiple outline lengths; the gains
also translate to downstream stories. All code, data, and models are
open-sourced. | cs.CL | Improving Pacing in Long-Form Story Planning |
2023-11-08T04:34:30Z | http://arxiv.org/pdf/2311.04453v2 | http://arxiv.org/abs/2311.04453v2 | Ryo Ueda, Tadahiro Taniguchi | 2024-04-02T07:53:07Z | cs.CL | As a sub-discipline of evolutionary and computational linguistics, emergent
communication (EC) studies communication protocols, called emergent languages,
arising in simulations where agents communicate. A key goal of EC is to give
rise to languages that share statistical properties with natural languages. In
this paper, we reinterpret Lewis's signaling game, a frequently used setting in
EC, as beta-VAE and reformulate its objective function as ELBO. Consequently,
we clarify the existence of prior distributions of emergent languages and show
that the choice of the priors can influence their statistical properties.
Specifically, we address the properties of word lengths and segmentation, known
as Zipf's law of abbreviation (ZLA) and Harris's articulation scheme (HAS),
respectively. It has been reported that the emergent languages do not follow
them when using the conventional objective. We experimentally demonstrate that
by selecting an appropriate prior distribution, more natural segments emerge,
while suggesting that the conventional one prevents the languages from
following ZLA and HAS. | cs.CL | Lewis's Signaling Game as beta-VAE For Natural Word Lengths and Segments |
2023-11-08T04:20:56Z | http://arxiv.org/pdf/2311.04449v1 | http://arxiv.org/abs/2311.04449v1 | Jishnu Ray Chowdhury, Cornelia Caragea | 2023-11-08T04:20:56Z | cs.LG, cs.CL | Binary Balanced Tree RvNNs (BBT-RvNNs) enforce sequence composition according
to a preset balanced binary tree structure. Thus, their non-linear recursion
depth is just $\log_2 n$ ($n$ being the sequence length). Such logarithmic
scaling makes BBT-RvNNs efficient and scalable on long sequence tasks such as
Long Range Arena (LRA). However, such computational efficiency comes at a cost
because BBT-RvNNs cannot solve simple arithmetic tasks like ListOps. On the
flip side, RvNNs (e.g., Beam Tree RvNN) that do succeed on ListOps (and other
structure-sensitive tasks like formal logical inference) are generally several
times more expensive than even RNNs. In this paper, we introduce a novel
framework -- Recursion in Recursion (RIR) to strike a balance between the two
sides - getting some of the benefits from both worlds. In RIR, we use a form of
two-level nested recursion - where the outer recursion is a $k$-ary balanced
tree model with another recursive model (inner recursion) implementing its cell
function. For the inner recursion, we choose Beam Tree RvNNs (BT-RvNN). To
adjust BT-RvNNs within RIR we also propose a novel strategy of beam alignment.
Overall, this entails that the total recursive depth in RIR is upper-bounded by
$k \log_k n$. Our best RIR-based model is the first model that demonstrates
high ($\geq 90\%$) length-generalization performance on ListOps while at the
same time being scalable enough to be trainable on long sequence inputs from
LRA. Moreover, in terms of accuracy in the LRA language tasks, it performs
competitively with Structured State Space Models (SSMs) without any special
initialization - outperforming Transformers by a large margin. On the other
hand, while SSMs can marginally outperform RIR on LRA, they (SSMs) fail to
length-generalize on ListOps. Our code is available at:
\url{https://github.com/JRC1995/BeamRecursionFamily/}. | cs.LG | Recursion in Recursion: Two-Level Nested Recursion for Length Generalization with Scalability |
2023-11-08T01:45:37Z | http://arxiv.org/pdf/2311.04939v1 | http://arxiv.org/abs/2311.04939v1 | Jiaqi Li, Mengmeng Wang, Zilong Zheng, Muhan Zhang | 2023-11-08T01:45:37Z | cs.CL, cs.AI | Large language models (LLMs), despite their impressive performance in various
language tasks, are typically limited to processing texts within context-window
size. This limitation has spurred significant research efforts to enhance LLMs'
long-context understanding with high-quality long-sequence benchmarks. However,
prior datasets in this regard suffer from shortcomings, such as short context
length compared to the context window of modern LLMs; outdated documents that
have data leakage problems; and an emphasis on short dependency tasks rather
than long dependency tasks. In this paper, we present LooGLE, a Long Context
Generic Language Evaluation benchmark for LLMs' long context understanding.
LooGLE features relatively new documents post-2022, with over 24,000 tokens per
document and 6,000 newly generated questions spanning diverse domains. Human
annotators meticulously crafted more than 1,100 high-quality question-answer
pairs to meet the long dependency requirements. These pairs underwent thorough
cross-validation, yielding the most precise assessment of LLMs' long dependency
capabilities. The evaluation of eight state-of-the-art LLMs on LooGLE revealed
key findings: (i) commercial models outperformed open-sourced models; (ii) LLMs
excelled in short dependency tasks like short question-answering and cloze
tasks but struggled with more intricate long dependency tasks; (iii) in-context
learning and chaining thoughts offered only marginal improvements; (iv)
retrieval-based techniques demonstrated substantial benefits for short
question-answering, while strategies for extending context window length had
limited impact on long context understanding. As such, LooGLE not only provides
a systematic and comprehensive evaluation schema on long-context LLMs, but also
sheds light on future development of enhanced models towards "true long-context
understanding". | cs.CL | LooGLE: Can Long-Context Language Models Understand Long Contexts? |
2023-11-08T01:27:34Z | http://arxiv.org/pdf/2311.04420v1 | http://arxiv.org/abs/2311.04420v1 | Xiang Zhou, Yichen Jiang, Mohit Bansal | 2023-11-08T01:27:34Z | cs.CL, cs.AI, cs.LG | Recent diagnostic datasets on compositional generalization, such as SCAN
(Lake and Baroni, 2018) and COGS (Kim and Linzen, 2020), expose severe problems
in models trained from scratch on these datasets. However, in contrast to this
poor performance, state-of-the-art models trained on larger and more general
datasets show better generalization ability. In this work, to reconcile this
inconsistency, we conduct an empirical analysis by training Transformer models
on a variety of training sets with different data factors, including dataset
scale, pattern complexity, example difficulty, etc. First, we show that
increased dataset complexity can lead to better generalization behavior on
multiple different generalization challenges. To further understand this
improvement, we show two axes of the benefit from more complex datasets: they
provide more diverse examples so compositional understanding becomes more
effective, and they also prevent ungeneralizable memorization of the examples
due to reduced example repetition frequency. Finally, we explore how training
examples of different difficulty levels influence generalization differently.
On synthetic datasets, simple examples invoke stronger compositionality than
hard examples do. On larger-scale real language datasets, while hard examples
become more important potentially to ensure decent data coverage, a balanced
mixture of simple and hard examples manages to induce the strongest
generalizability. The code and data for this work are available at
https://github.com/owenzx/data4comp | cs.CL | Data Factors for Better Compositional Generalization |
2023-11-07T22:52:54Z | http://arxiv.org/pdf/2311.04378v2 | http://arxiv.org/abs/2311.04378v2 | Hanlin Zhang, Benjamin L. Edelman, Danilo Francati, Daniele Venturi, Giuseppe Ateniese, Boaz Barak | 2023-11-15T00:21:29Z | cs.LG, cs.CL, cs.CR | Watermarking generative models consists of planting a statistical signal
(watermark) in a model's output so that it can be later verified that the
output was generated by the given model. A strong watermarking scheme satisfies
the property that a computationally bounded attacker cannot erase the watermark
without causing significant quality degradation. In this paper, we study the
(im)possibility of strong watermarking schemes. We prove that, under
well-specified and natural assumptions, strong watermarking is impossible to
achieve. This holds even in the private detection algorithm setting, where the
watermark insertion and detection algorithms share a secret key, unknown to the
attacker. To prove this result, we introduce a generic efficient watermark
attack; the attacker is not required to know the private key of the scheme or
even which scheme is used. Our attack is based on two assumptions: (1) The
attacker has access to a "quality oracle" that can evaluate whether a candidate
output is a high-quality response to a prompt, and (2) The attacker has access
to a "perturbation oracle" which can modify an output with a nontrivial
probability of maintaining quality, and which induces an efficiently mixing
random walk on high-quality outputs. We argue that both assumptions can be
satisfied in practice by an attacker with weaker computational capabilities
than the watermarked model itself, to which the attacker has only black-box
access. Furthermore, our assumptions will likely only be easier to satisfy over
time as models grow in capabilities and modalities. We demonstrate the
feasibility of our attack by instantiating it to attack three existing
watermarking schemes for large language models: Kirchenbauer et al. (2023),
Kuditipudi et al. (2023), and Zhao et al. (2023). The same attack successfully
removes the watermarks planted by all three schemes, with only minor quality
degradation. | cs.LG | Watermarks in the Sand: Impossibility of Strong Watermarking for Generative Models |
2023-11-07T22:23:51Z | http://arxiv.org/pdf/2311.04368v1 | http://arxiv.org/abs/2311.04368v1 | Jason Holmes, Rui Peng, Yiwei Li, Jinyu Hu, Zhengliang Liu, Zihao Wu, Huan Zhao, Xi Jiang, Wei Liu, Hong Wei, Jie Zou, Tianming Liu, Yi Shao | 2023-11-07T22:23:51Z | cs.CL | IMPORTANCE The response effectiveness of different large language models
(LLMs) and various individuals, including medical students, graduate students,
and practicing physicians, in pediatric ophthalmology consultations, has not
been clearly established yet. OBJECTIVE Design a 100-question exam based on
pediatric ophthalmology to evaluate the performance of LLMs in highly
specialized scenarios and compare them with the performance of medical students
and physicians at different levels. DESIGN, SETTING, AND PARTICIPANTS This
survey study assessed three LLMs, namely ChatGPT (GPT-3.5), GPT-4, and PaLM2,
were assessed alongside three human cohorts: medical students, postgraduate
students, and attending physicians, in their ability to answer questions
related to pediatric ophthalmology. It was conducted by administering
questionnaires in the form of test papers through the LLM network interface,
with the valuable participation of volunteers. MAIN OUTCOMES AND MEASURES Mean
scores of LLM and humans on 100 multiple-choice questions, as well as the
answer stability, correlation, and response confidence of each LLM. RESULTS
GPT-4 performed comparably to attending physicians, while ChatGPT (GPT-3.5) and
PaLM2 outperformed medical students but slightly trailed behind postgraduate
students. Furthermore, GPT-4 exhibited greater stability and confidence when
responding to inquiries compared to ChatGPT (GPT-3.5) and PaLM2. CONCLUSIONS
AND RELEVANCE Our results underscore the potential for LLMs to provide medical
assistance in pediatric ophthalmology and suggest significant capacity to guide
the education of medical students. | cs.CL | Evaluating multiple large language models in pediatric ophthalmology |
2023-11-07T21:59:16Z | http://arxiv.org/pdf/2311.04364v1 | http://arxiv.org/abs/2311.04364v1 | Danial Kamali, Parisa Kordjamshidi | 2023-11-07T21:59:16Z | cs.CL, cs.RO | Compositional generalization, the ability of intelligent models to
extrapolate understanding of components to novel compositions, is a fundamental
yet challenging facet in AI research, especially within multimodal
environments. In this work, we address this challenge by exploiting the
syntactic structure of language to boost compositional generalization. This
paper elevates the importance of syntactic grounding, particularly through
attention masking techniques derived from text input parsing. We introduce and
evaluate the merits of using syntactic information in the multimodal grounding
problem. Our results on grounded compositional generalization underscore the
positive impact of dependency parsing across diverse tasks when utilized with
Weight Sharing across the Transformer encoder. The results push the
state-of-the-art in multimodal grounding and parameter-efficient modeling and
provide insights for future research. | cs.CL | Syntax-Guided Transformers: Elevating Compositional Generalization and Grounding in Multimodal Environments |
2023-11-07T21:27:17Z | http://arxiv.org/pdf/2311.04354v2 | http://arxiv.org/abs/2311.04354v2 | Michael A. Lepori, Thomas Serre, Ellie Pavlick | 2023-11-17T15:15:17Z | cs.CL | Neural network models have achieved high performance on a wide variety of
complex tasks, but the algorithms that they implement are notoriously difficult
to interpret. In order to understand these algorithms, it is often necessary to
hypothesize intermediate variables involved in the network's computation. For
example, does a language model depend on particular syntactic properties when
generating a sentence? However, existing analysis tools make it difficult to
test hypotheses of this type. We propose a new analysis technique -- circuit
probing -- that automatically uncovers low-level circuits that compute
hypothesized intermediate variables. This enables causal analysis through
targeted ablation at the level of model parameters. We apply this method to
models trained on simple arithmetic tasks, demonstrating its effectiveness at
(1) deciphering the algorithms that models have learned, (2) revealing modular
structure within a model, and (3) tracking the development of circuits over
training. We compare circuit probing to other methods across these three
experiments, and find it on par or more effective than existing analysis
methods. Finally, we demonstrate circuit probing on a real-world use case,
uncovering circuits that are responsible for subject-verb agreement and
reflexive anaphora in GPT2-Small and Medium. | cs.CL | Uncovering Intermediate Variables in Transformers using Circuit Probing |
2023-11-07T21:09:57Z | http://arxiv.org/pdf/2311.04348v1 | http://arxiv.org/abs/2311.04348v1 | Sai Munikoti, Anurag Acharya, Sridevi Wagle, Sameera Horawalavithana | 2023-11-07T21:09:57Z | cs.CL, cs.AI, I.2.7 | Despite the dramatic progress in Large Language Model (LLM) development, LLMs
often provide seemingly plausible but not factual information, often referred
to as hallucinations. Retrieval-augmented LLMs provide a non-parametric
approach to solve these issues by retrieving relevant information from external
data sources and augment the training process. These models help to trace
evidence from an externally provided knowledge base allowing the model
predictions to be better interpreted and verified. In this work, we critically
evaluate these models in their ability to perform in scientific document
reasoning tasks. To this end, we tuned multiple such model variants with
science-focused instructions and evaluated them on a scientific document
reasoning benchmark for the usefulness of the retrieved document passages. Our
findings suggest that models justify predictions in science tasks with
fabricated evidence and leveraging scientific corpus as pretraining data does
not alleviate the risk of evidence fabrication. | cs.CL | Evaluating the Effectiveness of Retrieval-Augmented Large Language Models in Scientific Document Reasoning |
2023-11-07T21:00:51Z | http://arxiv.org/pdf/2311.04345v1 | http://arxiv.org/abs/2311.04345v1 | Wenbo Zhang, Hangzhi Guo, Ian D Kivlichan, Vinodkumar Prabhakaran, Davis Yadav, Amulya Yadav | 2023-11-07T21:00:51Z | cs.CL, cs.AI | Toxicity is an increasingly common and severe issue in online spaces.
Consequently, a rich line of machine learning research over the past decade has
focused on computationally detecting and mitigating online toxicity. These
efforts crucially rely on human-annotated datasets that identify toxic content
of various kinds in social media texts. However, such annotations historically
yield low inter-rater agreement, which was often dealt with by taking the
majority vote or other such approaches to arrive at a single ground truth
label. Recent research has pointed out the importance of accounting for the
subjective nature of this task when building and utilizing these datasets, and
this has triggered work on analyzing and better understanding rater
disagreements, and how they could be effectively incorporated into the machine
learning developmental pipeline. While these efforts are filling an important
gap, there is a lack of a broader framework about the root causes of rater
disagreement, and therefore, we situate this work within that broader
landscape. In this survey paper, we analyze a broad set of literature on the
reasons behind rater disagreements focusing on online toxicity, and propose a
detailed taxonomy for the same. Further, we summarize and discuss the potential
solutions targeting each reason for disagreement. We also discuss several open
issues, which could promote the future development of online toxicity research. | cs.CL | A Taxonomy of Rater Disagreements: Surveying Challenges & Opportunities from the Perspective of Annotating Online Toxicity |
2023-11-07T20:38:30Z | http://arxiv.org/pdf/2311.04335v1 | http://arxiv.org/abs/2311.04335v1 | Sihao Chen, Hongming Zhang, Tong Chen, Ben Zhou, Wenhao Yu, Dian Yu, Baolin Peng, Hongwei Wang, Dan Roth, Dong Yu | 2023-11-07T20:38:30Z | cs.CL, cs.AI | We introduce sub-sentence encoder, a contrastively-learned contextual
embedding model for fine-grained semantic representation of text. In contrast
to the standard practice with sentence embeddings, where the meaning of an
entire sequence of text is encoded into a fixed-length vector, the sub-sentence
encoder learns to produce distinct contextual embeddings corresponding to
different atomic propositions, i.e. atomic units of meaning expressed within a
text sequence. The sub-sentence embeddings are contrastively learned to
recognize (inferred) semantic equivalence between propositions across different
text sequences. Our experiments show the effectiveness of sub-sentence encoders
in applications, such as retrieving supporting facts for fine-grained text
attribution or recognizing the conditional semantic similarity between texts.
In practice, we demonstrate that sub-sentence encoders keep the same level of
inference cost and space complexity compared to sentence encoders. | cs.CL | Sub-Sentence Encoder: Contrastive Learning of Propositional Semantic Representations |
2023-11-07T20:21:42Z | http://arxiv.org/pdf/2311.04329v2 | http://arxiv.org/abs/2311.04329v2 | Ryan Cotterell, Anej Svete, Clara Meister, Tianyu Liu, Li Du | 2024-04-17T07:31:01Z | cs.CL | Large language models have become one of the most commonly deployed NLP
inventions. In the past half-decade, their integration into core natural
language processing tools has dramatically increased the performance of such
tools, and they have entered the public discourse surrounding artificial
intelligence. Consequently, it is important for both developers and researchers
alike to understand the mathematical foundations of large language models, as
well as how to implement them. These notes are the accompaniment to the
theoretical portion of the ETH Z\"urich course on large language models,
covering what constitutes a language model from a formal, theoretical
perspective. | cs.CL | Formal Aspects of Language Modeling |
2023-11-07T19:32:48Z | http://arxiv.org/pdf/2311.04936v1 | http://arxiv.org/abs/2311.04936v1 | Andrei Barcovschi, Rishabh Jain, Peter Corcoran | 2023-11-07T19:32:48Z | cs.CL, cs.AI, cs.SD, eess.AS | Automatic Speech Recognition (ASR) systems have progressed significantly in
their performance on adult speech data; however, transcribing child speech
remains challenging due to the acoustic differences in the characteristics of
child and adult voices. This work aims to explore the potential of adapting
state-of-the-art Conformer-transducer models to child speech to improve child
speech recognition performance. Furthermore, the results are compared with
those of self-supervised wav2vec2 models and semi-supervised multi-domain
Whisper models that were previously finetuned on the same data. We demonstrate
that finetuning Conformer-transducer models on child speech yields significant
improvements in ASR performance on child speech, compared to the non-finetuned
models. We also show Whisper and wav2vec2 adaptation on different child speech
datasets. Our detailed comparative analysis shows that wav2vec2 provides the
most consistent performance improvements among the three methods studied. | cs.CL | A comparative analysis between Conformer-Transducer, Whisper, and wav2vec2 for improving the child speech recognition |
2023-11-07T19:06:31Z | http://arxiv.org/pdf/2311.04292v1 | http://arxiv.org/abs/2311.04292v1 | Zhongfen Deng, Seunghyun Yoon, Trung Bui, Franck Dernoncourt, Quan Hung Tran, Shuaiqi Liu, Wenting Zhao, Tao Zhang, Yibo Wang, Philip S. Yu | 2023-11-07T19:06:31Z | cs.CL | Aspect-based meeting transcript summarization aims to produce multiple
summaries, each focusing on one aspect of content in a meeting transcript. It
is challenging as sentences related to different aspects can mingle together,
and those relevant to a specific aspect can be scattered throughout the long
transcript of a meeting. The traditional summarization methods produce one
summary mixing information of all aspects, which cannot deal with the above
challenges of aspect-based meeting transcript summarization. In this paper, we
propose a two-stage method for aspect-based meeting transcript summarization.
To select the input content related to specific aspects, we train a sentence
classifier on a dataset constructed from the AMI corpus with pseudo-labeling.
Then we merge the sentences selected for a specific aspect as the input for the
summarizer to produce the aspect-based summary. Experimental results on the AMI
corpus outperform many strong baselines, which verifies the effectiveness of
our proposed method. | cs.CL | Aspect-based Meeting Transcript Summarization: A Two-Stage Approach with Weak Supervision on Sentence Classification |
2023-11-07T19:00:44Z | http://arxiv.org/pdf/2311.04284v1 | http://arxiv.org/abs/2311.04284v1 | Angelika Romanou, Syrielle Montariol, Debjit Paul, Leo Laugier, Karl Aberer, Antoine Bosselut | 2023-11-07T19:00:44Z | cs.CL, cs.AI | Understanding narratives requires reasoning about the cause-and-effect
relationships between events mentioned in the text. While existing foundation
models yield impressive results in many NLP tasks requiring reasoning, it is
unclear whether they understand the complexity of the underlying network of
causal relationships of events in narratives. In this work, we present CRAB, a
new Causal Reasoning Assessment Benchmark designed to evaluate causal
understanding of events in real-world narratives. CRAB contains fine-grained,
contextual causality annotations for ~2.7K pairs of real-world events that
describe various newsworthy event timelines (e.g., the acquisition of Twitter
by Elon Musk). Using CRAB, we measure the performance of several large language
models, demonstrating that most systems achieve poor performance on the task.
Motivated by classical causal principles, we also analyze the causal structures
of groups of events in CRAB, and find that models perform worse on causal
reasoning when events are derived from complex causal structures compared to
simple linear causal chains. We make our dataset and code available to the
research community. | cs.CL | CRAB: Assessing the Strength of Causal Relationships Between Real-world Events |
2023-11-07T18:43:34Z | http://arxiv.org/pdf/2311.04205v2 | http://arxiv.org/abs/2311.04205v2 | Yihe Deng, Weitong Zhang, Zixiang Chen, Quanquan Gu | 2024-04-18T23:49:56Z | cs.CL, cs.AI, cs.LG | Misunderstandings arise not only in interpersonal communication but also
between humans and Large Language Models (LLMs). Such discrepancies can make
LLMs interpret seemingly unambiguous questions in unexpected ways, yielding
incorrect responses. While it is widely acknowledged that the quality of a
prompt, such as a question, significantly impacts the quality of the response
provided by LLMs, a systematic method for crafting questions that LLMs can
better comprehend is still underdeveloped. In this paper, we present a method
named `Rephrase and Respond' (RaR), which allows LLMs to rephrase and expand
questions posed by humans and provide responses in a single prompt. This
approach serves as a simple yet effective prompting method for improving
performance. We also introduce a two-step variant of RaR, where a rephrasing
LLM first rephrases the question and then passes the original and rephrased
questions together to a different responding LLM. This facilitates the
effective utilization of rephrased questions generated by one LLM with another.
Our experiments demonstrate that our methods significantly improve the
performance of different models across a wide range to tasks. We further
provide a comprehensive comparison between RaR and the popular Chain-of-Thought
(CoT) methods, both theoretically and empirically. We show that RaR is
complementary to CoT and can be combined with CoT to achieve even better
performance. Our work not only contributes to enhancing LLM performance
efficiently and effectively but also sheds light on a fair evaluation of LLM
capabilities. Data and codes are available at
https://github.com/uclaml/Rephrase-and-Respond. | cs.CL | Rephrase and Respond: Let Large Language Models Ask Better Questions for Themselves |
2023-11-07T18:39:10Z | http://arxiv.org/pdf/2311.04199v1 | http://arxiv.org/abs/2311.04199v1 | Peilin Zhou, Meng Cao, You-Liang Huang, Qichen Ye, Peiyan Zhang, Junling Liu, Yueqi Xie, Yining Hua, Jaeboum Kim | 2023-11-07T18:39:10Z | cs.IR, cs.CL | Large Multimodal Models (LMMs) have demonstrated impressive performance
across various vision and language tasks, yet their potential applications in
recommendation tasks with visual assistance remain unexplored. To bridge this
gap, we present a preliminary case study investigating the recommendation
capabilities of GPT-4V(ison), a recently released LMM by OpenAI. We construct a
series of qualitative test samples spanning multiple domains and employ these
samples to assess the quality of GPT-4V's responses within recommendation
scenarios. Evaluation results on these test samples prove that GPT-4V has
remarkable zero-shot recommendation abilities across diverse domains, thanks to
its robust visual-text comprehension capabilities and extensive general
knowledge. However, we have also identified some limitations in using GPT-4V
for recommendations, including a tendency to provide similar responses when
given similar inputs. This report concludes with an in-depth discussion of the
challenges and research opportunities associated with utilizing GPT-4V in
recommendation scenarios. Our objective is to explore the potential of
extending LMMs from vision and language tasks to recommendation tasks. We hope
to inspire further research into next-generation multimodal generative
recommendation models, which can enhance user experiences by offering greater
diversity and interactivity. All images and prompts used in this report will be
accessible at https://github.com/PALIN2018/Evaluate_GPT-4V_Rec. | cs.IR | Exploring Recommendation Capabilities of GPT-4V(ision): A Preliminary Case Study |
2023-11-07T18:36:16Z | http://arxiv.org/pdf/2311.04196v1 | http://arxiv.org/abs/2311.04196v1 | Zhongfen Deng, Hao Peng, Tao Zhang, Shuaiqi Liu, Wenting Zhao, Yibo Wang, Philip S. Yu | 2023-11-07T18:36:16Z | cs.CL, cs.AI | Product attribute value extraction is an important task in e-Commerce which
can help several downstream applications such as product search and
recommendation. Most previous models handle this task using sequence labeling
or question answering method which rely on the sequential position information
of values in the product text and are vulnerable to data discrepancy between
training and testing. This limits their generalization ability to real-world
scenario in which each product can have multiple descriptions across various
shopping platforms with different composition of text and style. They also have
limited zero-shot ability to new values. In this paper, we propose a multi-task
learning model with value generation/classification and attribute prediction
called JPAVE to predict values without the necessity of position information of
values in the text. Furthermore, the copy mechanism in value generator and the
value attention module in value classifier help our model address the data
discrepancy issue by only focusing on the relevant part of input text and
ignoring other information which causes the discrepancy issue such as sentence
structure in the text. Besides, two variants of our model are designed for
open-world and closed-world scenarios. In addition, copy mechanism introduced
in the first variant based on value generation can improve its zero-shot
ability for identifying unseen values. Experimental results on a public dataset
demonstrate the superiority of our model compared with strong baselines and its
generalization ability of predicting new values. | cs.CL | JPAVE: A Generation and Classification-based Model for Joint Product Attribute Prediction and Value Extraction |
2023-11-07T18:33:34Z | http://arxiv.org/pdf/2311.04192v1 | http://arxiv.org/abs/2311.04192v1 | Yuiga Wada, Kanta Kaneda, Komei Sugiura | 2023-11-07T18:33:34Z | cs.CV, cs.CL | Image captioning studies heavily rely on automatic evaluation metrics such as
BLEU and METEOR. However, such n-gram-based metrics have been shown to
correlate poorly with human evaluation, leading to the proposal of alternative
metrics such as SPICE for English; however, no equivalent metrics have been
established for other languages. Therefore, in this study, we propose an
automatic evaluation metric called JaSPICE, which evaluates Japanese captions
based on scene graphs. The proposed method generates a scene graph from
dependencies and the predicate-argument structure, and extends the graph using
synonyms. We conducted experiments employing 10 image captioning models trained
on STAIR Captions and PFN-PIC and constructed the Shichimi dataset, which
contains 103,170 human evaluations. The results showed that our metric
outperformed the baseline metrics for the correlation coefficient with the
human evaluation. | cs.CV | JaSPICE: Automatic Evaluation Metric Using Predicate-Argument Structures for Image Captioning Models |
2023-11-07T18:32:34Z | http://arxiv.org/pdf/2311.04189v1 | http://arxiv.org/abs/2311.04189v1 | Yevhen Kostiuk, Grigori Sidorov, Olga Kolesnikova | 2023-11-07T18:32:34Z | cs.CL | In natural language processing (NLP), lexical function is a concept to
unambiguously represent semantic and syntactic features of words and phrases in
text first crafted in the Meaning-Text Theory. Hierarchical classification of
lexical functions involves organizing these features into a tree-like hierarchy
of categories or labels. This is a challenging task as it requires a good
understanding of the context and the relationships among words and phrases in
text. It also needs large amounts of labeled data to train language models
effectively. In this paper, we present a dataset of most frequent Spanish
verb-noun collocations and sentences where they occur, each collocation is
assigned to one of 37 lexical functions defined as classes for a hierarchical
classification task. Each class represents a relation between the noun and the
verb in a collocation involving their semantic and syntactic features. We
combine the classes in a tree-based structure, and introduce classification
objectives for each level of the structure. The dataset was created by
dependency tree parsing and matching of the phrases in Spanish news. We provide
baselines and data splits for each objective. | cs.CL | SpaDeLeF: A Dataset for Hierarchical Classification of Lexical Functions for Collocations in Spanish |
2023-11-07T18:17:05Z | http://arxiv.org/pdf/2311.04934v2 | http://arxiv.org/abs/2311.04934v2 | In Gim, Guojun Chen, Seung-seob Lee, Nikhil Sarda, Anurag Khandelwal, Lin Zhong | 2024-04-25T15:45:19Z | cs.CL, cs.AI | We present Prompt Cache, an approach for accelerating inference for large
language models (LLM) by reusing attention states across different LLM prompts.
Many input prompts have overlapping text segments, such as system messages,
prompt templates, and documents provided for context. Our key insight is that
by precomputing and storing the attention states of these frequently occurring
text segments on the inference server, we can efficiently reuse them when these
segments appear in user prompts. Prompt Cache employs a schema to explicitly
define such reusable text segments, called prompt modules. The schema ensures
positional accuracy during attention state reuse and provides users with an
interface to access cached states in their prompt. Using a prototype
implementation, we evaluate Prompt Cache across several LLMs. We show that
Prompt Cache significantly reduce latency in time-to-first-token, especially
for longer prompts such as document-based question answering and
recommendations. The improvements range from 8x for GPU-based inference to 60x
for CPU-based inference, all while maintaining output accuracy and without the
need for model parameter modifications. | cs.CL | Prompt Cache: Modular Attention Reuse for Low-Latency Inference |
2023-11-07T18:03:23Z | http://arxiv.org/pdf/2311.04177v1 | http://arxiv.org/abs/2311.04177v1 | Eric Melz | 2023-11-07T18:03:23Z | cs.CL, cs.AI | Large Language Models (LLMs) are smart but forgetful. Recent studies, (e.g.,
(Bubeck et al., 2023)) on modern LLMs have shown that they are capable of
performing amazing tasks typically necessitating human-level intelligence.
However, unlike humans, frozen LLMs do not improve over time; they neither
acquire new knowledge nor learn from their successes or failures. Some
approaches to improving the intelligence of LLMs include fine-tuning models
based on problem-solving performance (Zelikman et al., 2022), and building
bigger and more sophisticated models (Bubeck et al., 2023). However, these
methods have the drawback of requiring substantial data and computational
resources to retrain existing models. In this paper, we explore the use of
Retrieval Augmented Generation, also known as RAG (Lewis et al., 2021) to
improve problem-solving performance. We propose ARM-RAG (Auxiliary Rationale
Memory for Retrieval Augmented Generation), a system that learns from its
successes without incurring high training costs. We demonstrate that the
storage and subsequent retrieval of reasoning chains have a positive influence
on performance in grade-school math problems. | cs.CL | Enhancing LLM Intelligence with ARM-RAG: Auxiliary Rationale Memory for Retrieval Augmented Generation |
2023-11-07T17:48:35Z | http://arxiv.org/pdf/2311.04166v2 | http://arxiv.org/abs/2311.04166v2 | Ruchit Rawal, Mariya Toneva | 2024-06-14T18:36:36Z | cs.CL, cs.LG | The rapid growth in natural language processing (NLP) research has led to
numerous new models, outpacing our understanding of how they compare to
established ones. One major reason for this difficulty is saturating
benchmarks, which may not well reflect differences in model performance in the
wild. In this work, we introduce a novel framework to compare two NLP models by
revealing their shared invariance to interpretable input perturbations
targeting a specific linguistic capability. Via experiments on models from the
same and different architecture families, this framework offers insights about
how changes in models (e.g., distillation, size increase) affect linguistic
capabilities. Furthermore, our framework enables evaluation of invariances
between commercial black-box models (e.g., InstructGPT family) and models that
are better understood (e.g., GPT-2). Across experiments, we observe that large
language models share many invariances encoded by models of various sizes,
whereas the invariances by large models are only shared by other large models.
Possessing a wide variety of invariances may be key to the recent successes of
large language models, and our framework can shed light on the types of
invariances retained or emerging in new models. We make the code publicly
available. | cs.CL | Perturbed examples reveal invariances shared by language models |
2023-11-07T17:31:50Z | http://arxiv.org/pdf/2311.04155v3 | http://arxiv.org/abs/2311.04155v3 | Jiale Cheng, Xiao Liu, Kehan Zheng, Pei Ke, Hongning Wang, Yuxiao Dong, Jie Tang, Minlie Huang | 2024-06-21T06:06:07Z | cs.CL | Large language models (LLMs) have shown impressive success in various
applications. However, these models are often not well aligned with human
intents, which calls for additional treatments on them; that is, the alignment
problem. To make LLMs better follow user instructions, existing alignment
methods primarily focus on further training them. However, the extra training
of LLMs is usually expensive in terms of GPU computing; even worse, some LLMs
are not accessible for user-demanded training, such as GPTs. In this work, we
take a different perspective -- Black-Box Prompt Optimization (BPO) -- to
perform alignments. The idea is to optimize user prompts to suit LLMs' input
understanding, so as to best realize users' intents without updating LLMs'
parameters. BPO leverages human preferences to optimize prompts, thus making it
superior to LLM (e.g., ChatGPT) as a prompt engineer. Moreover, BPO is
model-agnostic, and the empirical results demonstrate that the BPO-aligned
ChatGPT yields a 22% increase in the win rate against its original version and
10% for GPT-4. Notably, the BPO-aligned LLMs can outperform the same models
aligned by PPO and DPO, and it also brings additional performance gains when
combining BPO with PPO or DPO. Code and datasets are released at
https://github.com/thu-coai/BPO. | cs.CL | Black-Box Prompt Optimization: Aligning Large Language Models without Model Training |
2023-11-07T17:13:40Z | http://arxiv.org/pdf/2311.04142v1 | http://arxiv.org/abs/2311.04142v1 | Manas Mohanty, Tanya Roosta, Peyman Passban | 2023-11-07T17:13:40Z | cs.CL | Deep neural networks (DNNs) have improved NLP tasks significantly, but
training and maintaining such networks could be costly. Model compression
techniques, such as, knowledge distillation (KD), have been proposed to address
the issue; however, the compression process could be lossy. Motivated by this,
our work investigates how a distilled student model differs from its teacher,
if the distillation process causes any information losses, and if the loss
follows a specific pattern. Our experiments aim to shed light on the type of
tasks might be less or more sensitive to KD by reporting data points on the
contribution of different factors, such as the number of layers or attention
heads. Results such as ours could be utilized when determining effective and
efficient configurations to achieve optimal information transfers between
larger (teacher) and smaller (student) models. | cs.CL | What is Lost in Knowledge Distillation? |
2023-11-07T17:12:39Z | http://arxiv.org/pdf/2311.04139v1 | http://arxiv.org/abs/2311.04139v1 | Guillem Senabre Prades | 2023-11-07T17:12:39Z | cs.CL | This paper provides different approaches for a binary sentiment
classification on a small training dataset. LLMs that provided state-of-the-art
results in sentiment analysis and similar domains are being used, such as BERT,
RoBERTa and XLNet. | cs.CL | Modelling Sentiment Analysis: LLMs and data augmentation techniques |
2023-11-07T16:58:51Z | http://arxiv.org/pdf/2311.04131v3 | http://arxiv.org/abs/2311.04131v3 | Michael Lan, Fazl Barez | 2024-02-22T17:07:39Z | cs.CL, cs.AI, cs.LG | While transformer models exhibit strong capabilities on linguistic tasks,
their complex architectures make them difficult to interpret. Recent work has
aimed to reverse engineer transformer models into human-readable
representations called circuits that implement algorithmic functions. We extend
this research by analyzing and comparing circuits for similar sequence
continuation tasks, which include increasing sequences of digits, number words,
and months. Through the application of circuit analysis techniques, we identify
key sub-circuits responsible for detecting sequence members and for predicting
the next member in a sequence. Our analysis reveals that semantically related
sequences rely on shared circuit subgraphs with analogous roles. Overall,
documenting shared computational structures enables better prediction of model
behaviors, identification of errors, and safer editing procedures. This
mechanistic understanding of transformers is a critical step towards building
more robust, aligned, and interpretable language models. | cs.CL | Interpreting Shared Circuits for Ordered Sequence Prediction in a Large Language Model |
2023-11-07T16:50:33Z | http://arxiv.org/pdf/2311.04124v1 | http://arxiv.org/abs/2311.04124v1 | George Kour, Marcel Zalmanovici, Naama Zwerdling, Esther Goldbraich, Ora Nova Fandina, Ateret Anaby-Tavor, Orna Raz, Eitan Farchi | 2023-11-07T16:50:33Z | cs.CL, cs.AI, cs.LG, I.2.7 | As large language models become more prevalent, their possible harmful or
inappropriate responses are a cause for concern. This paper introduces a unique
dataset containing adversarial examples in the form of questions, which we call
AttaQ, designed to provoke such harmful or inappropriate responses. We assess
the efficacy of our dataset by analyzing the vulnerabilities of various models
when subjected to it. Additionally, we introduce a novel automatic approach for
identifying and naming vulnerable semantic regions - input semantic areas for
which the model is likely to produce harmful outputs. This is achieved through
the application of specialized clustering techniques that consider both the
semantic similarity of the input attacks and the harmfulness of the model's
responses. Automatically identifying vulnerable semantic regions enhances the
evaluation of model weaknesses, facilitating targeted improvements to its
safety mechanisms and overall reliability. | cs.CL | Unveiling Safety Vulnerabilities of Large Language Models |
2023-11-07T16:19:45Z | http://arxiv.org/pdf/2311.04933v1 | http://arxiv.org/abs/2311.04933v1 | Jason Holmes, Shuyuan Ye, Yiwei Li, Shi-Nan Wu, Zhengliang Liu, Zihao Wu, Jinyu Hu, Huan Zhao, Xi Jiang, Wei Liu, Hong Wei, Jie Zou, Tianming Liu, Yi Shao | 2023-11-07T16:19:45Z | cs.CL, cs.AI | Purpose: The performance of three different large language models (LLMS)
(GPT-3.5, GPT-4, and PaLM2) in answering ophthalmology professional questions
was evaluated and compared with that of three different professional
populations (medical undergraduates, medical masters, and attending
physicians). Methods: A 100-item ophthalmology single-choice test was
administered to three different LLMs (GPT-3.5, GPT-4, and PaLM2) and three
different professional levels (medical undergraduates, medical masters, and
attending physicians), respectively. The performance of LLM was comprehensively
evaluated and compared with the human group in terms of average score,
stability, and confidence. Results: Each LLM outperformed undergraduates in
general, with GPT-3.5 and PaLM2 being slightly below the master's level, while
GPT-4 showed a level comparable to that of attending physicians. In addition,
GPT-4 showed significantly higher answer stability and confidence than GPT-3.5
and PaLM2. Conclusion: Our study shows that LLM represented by GPT-4 performs
better in the field of ophthalmology. With further improvements, LLM will bring
unexpected benefits in medical education and clinical decision making in the
near future. | cs.CL | Evaluating Large Language Models in Ophthalmology |
2023-11-07T15:56:19Z | http://arxiv.org/pdf/2311.04088v1 | http://arxiv.org/abs/2311.04088v1 | Semere Kiros Bitew, Vincent Schelstraete, Klim Zaporojets, Kimberly Van Nieuwenhove, Reitske Meganck, Chris Develder | 2023-11-07T15:56:19Z | cs.CL | In disentangling the heterogeneity observed in psychopathology, personality
of the patients is considered crucial. While it has been demonstrated that
personality traits are reflected in the language used by a patient, we
hypothesize that this enables automatic inference of the personality type
directly from speech utterances, potentially more accurately than through a
traditional questionnaire-based approach explicitly designed for personality
classification. To validate this hypothesis, we adopt natural language
processing (NLP) and standard machine learning tools for classification. We
test this on a dataset of recorded clinical diagnostic interviews (CDI) on a
sample of 79 patients diagnosed with major depressive disorder (MDD) -- a
condition for which differentiated treatment based on personality styles has
been advocated -- and classified into anaclitic and introjective personality
styles. We start by analyzing the interviews to see which linguistic features
are associated with each style, in order to gain a better understanding of the
styles. Then, we develop automatic classifiers based on (a) standardized
questionnaire responses; (b) basic text features, i.e., TF-IDF scores of words
and word sequences; (c) more advanced text features, using LIWC (linguistic
inquiry and word count) and context-aware features using BERT (bidirectional
encoder representations from transformers); (d) audio features. We find that
automated classification with language-derived features (i.e., based on LIWC)
significantly outperforms questionnaire-based classification models.
Furthermore, the best performance is achieved by combining LIWC with the
questionnaire features. This suggests that more work should be put into
developing linguistically based automated techniques for characterizing
personality, however questionnaires still to some extent complement such
methods. | cs.CL | Personality Style Recognition via Machine Learning: Identifying Anaclitic and Introjective Personality Styles from Patients' Speech |
2023-11-07T15:40:43Z | http://arxiv.org/pdf/2311.04076v5 | http://arxiv.org/abs/2311.04076v5 | Lindia Tjuatja, Valerie Chen, Sherry Tongshuang Wu, Ameet Talwalkar, Graham Neubig | 2024-02-06T04:16:17Z | cs.CL | As large language models (LLMs) become more capable, there is growing
excitement about the possibility of using LLMs as proxies for humans in
real-world tasks where subjective labels are desired, such as in surveys and
opinion polling. One widely-cited barrier to the adoption of LLMs as proxies
for humans in subjective tasks is their sensitivity to prompt wording - but
interestingly, humans also display sensitivities to instruction changes in the
form of response biases. We investigate the extent to which LLMs reflect human
response biases, if at all. We look to survey design, where human response
biases caused by changes in the wordings of "prompts" have been extensively
explored in social psychology literature. Drawing from these works, we design a
dataset and framework to evaluate whether LLMs exhibit human-like response
biases in survey questionnaires. Our comprehensive evaluation of nine models
shows that popular open and commercial LLMs generally fail to reflect
human-like behavior, particularly in models that have undergone RLHF.
Furthermore, even if a model shows a significant change in the same direction
as humans, we find that they are sensitive to perturbations that do not elicit
significant changes in humans. These results highlight the pitfalls of using
LLMs as human proxies, and underscore the need for finer-grained
characterizations of model behavior. Our code, dataset, and collected samples
are available at https://github.com/lindiatjuatja/BiasMonkey | cs.CL | Do LLMs exhibit human-like response biases? A case study in survey design |
2023-11-07T15:38:09Z | http://arxiv.org/pdf/2311.04260v1 | http://arxiv.org/abs/2311.04260v1 | Motonari Kambara, Komei Sugiura | 2023-11-07T15:38:09Z | cs.RO, cs.CL, cs.CV | This paper aims to develop a framework that enables a robot to execute tasks
based on visual information, in response to natural language instructions for
Fetch-and-Carry with Object Grounding (FCOG) tasks. Although there have been
many frameworks, they usually rely on manually given instruction sentences.
Therefore, evaluations have only been conducted with fixed tasks. Furthermore,
many multimodal language understanding models for the benchmarks only consider
discrete actions. To address the limitations, we propose a framework for the
full automation of the generation, execution, and evaluation of FCOG tasks. In
addition, we introduce an approach to solving the FCOG tasks by dividing them
into four distinct subtasks. | cs.RO | Fully Automated Task Management for Generation, Execution, and Evaluation: A Framework for Fetch-and-Carry Tasks with Natural Language Instructions in Continuous Space |
2023-11-07T15:36:40Z | http://arxiv.org/pdf/2311.04072v2 | http://arxiv.org/abs/2311.04072v2 | Geyang Guo, Ranchi Zhao, Tianyi Tang, Wayne Xin Zhao, Ji-Rong Wen | 2024-04-15T15:25:53Z | cs.CL | Alignment with human preference is a desired property of large language
models (LLMs). Currently, the main alignment approach is based on reinforcement
learning from human feedback (RLHF). Despite the effectiveness of RLHF, it is
intricate to implement and train, thus recent studies explore how to develop
alternative alignment approaches based on supervised fine-tuning (SFT). A major
limitation of SFT is that it essentially does imitation learning, which cannot
fully understand what are the expected behaviors. To address this issue, we
propose an improved alignment approach named FIGA. Different from prior
methods, we incorporate fine-grained (i.e., token or phrase level) quality
signals that are derived by contrasting good and bad responses. Our approach
has made two major contributions. Firstly, we curate a refined alignment
dataset that pairs initial responses and the corresponding revised ones.
Secondly, we devise a new loss function can leverage fine-grained quality
signals to instruct the learning of LLMs for alignment. Extensive experiments
have demonstrated the effectiveness of our approaches by comparing a number of
competitive baselines. | cs.CL | Beyond Imitation: Leveraging Fine-grained Quality Signals for Alignment |
2023-11-07T15:25:52Z | http://arxiv.org/pdf/2311.04064v2 | http://arxiv.org/abs/2311.04064v2 | Marc-Alexander Lutz, Bastian Schäfermeier, Rachael Sexton, Michael Sharp, Alden Dima, Stefan Faulstich, Jagan Mohini Aluri | 2023-12-06T17:17:00Z | cs.CL, cs.LG, I.2.7; I.7 | Maintenance work orders are commonly used to document information about wind
turbine operation and maintenance. This includes details about proactive and
reactive wind turbine downtimes, such as preventative and corrective
maintenance. However, the information contained in maintenance work orders is
often unstructured and difficult to analyze, presenting challenges for
decision-makers wishing to use it for optimizing operation and maintenance. To
address this issue, this work compares three different approaches to calculate
reliability by performance indicators from maintenance work orders. The first
approach involves manual labeling of the maintenance work orders by domain
experts, using the schema defined in an industrial guideline to assign the
label accordingly. The second approach involves the development of a model that
automatically labels the maintenance work orders using text classification
methods. Through this method, we are able to achieve macro average and weighted
average F1-Scores of 0.75 and 0.85 respectively. The third technique uses an
AI-assisted tagging tool to tag and structure the raw maintenance information,
together with a novel rule-based approach for extracting relevant maintenance
work orders for failure rate calculation. In our experiments the AI-assisted
tool leads to a 88% drop in tagging time in comparison to the other two
approaches, while expert labeling and text classification are more accurate in
KPI extraction. Overall, our findings make extracting maintenance information
from maintenance work orders more efficient, enable the assessment of
reliability key performance indicators and therefore support the optimization
of wind turbine operation and maintenance. | cs.CL | KPI Extraction from Maintenance Work Orders -- A Comparison of Expert Labeling, Text Classification and AI-Assisted Tagging for Computing Failure Rates of Wind Turbines |
2023-11-07T15:00:39Z | http://arxiv.org/pdf/2311.04046v1 | http://arxiv.org/abs/2311.04046v1 | Diogo Cruz, Edoardo Pona, Alex Holness-Tofts, Elias Schmied, Víctor Abia Alonso, Charlie Griffin, Bogdan-Ionut Cirstea | 2023-11-07T15:00:39Z | cs.LG, cs.CL | Many capable large language models (LLMs) are developed via self-supervised
pre-training followed by a reinforcement-learning fine-tuning phase, often
based on human or AI feedback. During this stage, models may be guided by their
inductive biases to rely on simpler features which may be easier to extract, at
a cost to robustness and generalisation. We investigate whether principles
governing inductive biases in the supervised fine-tuning of LLMs also apply
when the fine-tuning process uses reinforcement learning. Following Lovering et
al (2021), we test two hypotheses: that features more $\textit{extractable}$
after pre-training are more likely to be utilised by the final policy, and that
the evidence for/against a feature predicts whether it will be utilised.
Through controlled experiments on synthetic and natural language tasks, we find
statistically significant correlations which constitute strong evidence for
these hypotheses. | cs.LG | Reinforcement Learning Fine-tuning of Language Models is Biased Towards More Extractable Features |
2023-11-07T14:55:52Z | http://arxiv.org/pdf/2311.04044v3 | http://arxiv.org/abs/2311.04044v3 | Haoran Li, Dadi Guo, Donghao Li, Wei Fan, Qi Hu, Xin Liu, Chunkit Chan, Duanyi Yao, Yuan Yao, Yangqiu Song | 2024-06-01T14:04:21Z | cs.CL, cs.CR | The rapid development of language models (LMs) brings unprecedented
accessibility and usage for both models and users. On the one hand, powerful
LMs achieve state-of-the-art performance over numerous downstream NLP tasks. On
the other hand, more and more attention is paid to unrestricted model accesses
that may bring malicious privacy risks of data leakage. To address these
issues, many recent works propose privacy-preserving language models (PPLMs)
with differential privacy (DP). Unfortunately, different DP implementations
make it challenging for a fair comparison among existing PPLMs. In this paper,
we present PrivLM-Bench, a multi-perspective privacy evaluation benchmark to
empirically and intuitively quantify the privacy leakage of LMs. Instead of
only reporting DP parameters, PrivLM-Bench sheds light on the neglected
inference data privacy during actual usage. PrivLM-Bench first clearly defines
multi-faceted privacy objectives. Then, PrivLM-Bench constructs a unified
pipeline to perform private fine-tuning. Lastly, PrivLM-Bench performs existing
privacy attacks on LMs with pre-defined privacy objectives as the empirical
evaluation results. The empirical attack results are used to fairly and
intuitively evaluate the privacy leakage of various PPLMs. We conduct extensive
experiments on three datasets of GLUE for mainstream LMs. | cs.CL | PrivLM-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models |
2023-11-07T14:21:29Z | http://arxiv.org/pdf/2311.04257v2 | http://arxiv.org/abs/2311.04257v2 | Qinghao Ye, Haiyang Xu, Jiabo Ye, Ming Yan, Anwen Hu, Haowei Liu, Qi Qian, Ji Zhang, Fei Huang, Jingren Zhou | 2023-11-09T01:56:51Z | cs.CL, cs.CV | Multi-modal Large Language Models (MLLMs) have demonstrated impressive
instruction abilities across various open-ended tasks. However, previous
methods primarily focus on enhancing multi-modal capabilities. In this work, we
introduce a versatile multi-modal large language model, mPLUG-Owl2, which
effectively leverages modality collaboration to improve performance in both
text and multi-modal tasks. mPLUG-Owl2 utilizes a modularized network design,
with the language decoder acting as a universal interface for managing
different modalities. Specifically, mPLUG-Owl2 incorporates shared functional
modules to facilitate modality collaboration and introduces a modality-adaptive
module that preserves modality-specific features. Extensive experiments reveal
that mPLUG-Owl2 is capable of generalizing both text tasks and multi-modal
tasks and achieving state-of-the-art performances with a single generic model.
Notably, mPLUG-Owl2 is the first MLLM model that demonstrates the modality
collaboration phenomenon in both pure-text and multi-modal scenarios, setting a
pioneering path in the development of future multi-modal foundation models. | cs.CL | mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaboration |
2023-11-07T14:18:03Z | http://arxiv.org/pdf/2311.04020v1 | http://arxiv.org/abs/2311.04020v1 | Tanzir Pial, Shahreen Salim, Charuta Pethe, Allen Kim, Steven Skiena | 2023-11-07T14:18:03Z | cs.CL | Novels are often adapted into feature films, but the differences between the
two media usually require dropping sections of the source text from the movie
script. Here we study this screen adaptation process by constructing narrative
alignments using the Smith-Waterman local alignment algorithm coupled with
SBERT embedding distance to quantify text similarity between scenes and book
units. We use these alignments to perform an automated analysis of 40
adaptations, revealing insights into the screenwriting process concerning (i)
faithfulness of adaptation, (ii) importance of dialog, (iii) preservation of
narrative order, and (iv) gender representation issues reflective of the
Bechdel test. | cs.CL | Analyzing Film Adaptation through Narrative Alignment |
2023-11-07T14:14:36Z | http://arxiv.org/pdf/2311.06303v1 | http://arxiv.org/abs/2311.06303v1 | Lei Zhang, Markus Stricker | 2023-11-07T14:14:36Z | cond-mat.mtrl-sci, cs.CL, physics.chem-ph, H.4; H.5; I.5; I.7; J.2 | MatNexus is a specialized software for the automated collection, processing,
and analysis of text from scientific articles. Through an integrated suite of
modules, the MatNexus facilitates the retrieval of scientific articles,
processes textual data for insights, generates vector representations suitable
for machine learning, and offers visualization capabilities for word
embeddings. With the vast volume of scientific publications, MatNexus stands
out as an end-to-end tool for researchers aiming to gain insights from
scientific literature in material science, making the exploration of materials,
such as the electrocatalyst examples we show here, efficient and insightful. | cond-mat.mtrl-sci | MatNexus: A Comprehensive Text Mining and Analysis Suite for Materials Discover |
2023-11-07T14:11:10Z | http://arxiv.org/pdf/2311.14693v1 | http://arxiv.org/abs/2311.14693v1 | Munmun De Choudhury, Sachin R. Pendse, Neha Kumar | 2023-11-07T14:11:10Z | cs.CL, cs.AI, cs.CY, cs.HC | The past decade has been transformative for mental health research and
practice. The ability to harness large repositories of data, whether from
electronic health records (EHR), mobile devices, or social media, has revealed
a potential for valuable insights into patient experiences, promising early,
proactive interventions, as well as personalized treatment plans. Recent
developments in generative artificial intelligence, particularly large language
models (LLMs), show promise in leading digital mental health to uncharted
territory. Patients are arriving at doctors' appointments with information
sourced from chatbots, state-of-the-art LLMs are being incorporated in medical
software and EHR systems, and chatbots from an ever-increasing number of
startups promise to serve as AI companions, friends, and partners. This article
presents contemporary perspectives on the opportunities and risks posed by LLMs
in the design, development, and implementation of digital mental health tools.
We adopt an ecological framework and draw on the affordances offered by LLMs to
discuss four application areas -- care-seeking behaviors from individuals in
need of care, community care provision, institutional and medical care
provision, and larger care ecologies at the societal level. We engage in a
thoughtful consideration of whether and how LLM-based technologies could or
should be employed for enhancing mental health. The benefits and harms our
article surfaces could serve to help shape future research, advocacy, and
regulatory efforts focused on creating more responsible, user-friendly,
equitable, and secure LLM-based tools for mental health treatment and
intervention. | cs.CL | Benefits and Harms of Large Language Models in Digital Mental Health |
2023-11-07T13:54:01Z | http://arxiv.org/pdf/2311.03998v1 | http://arxiv.org/abs/2311.03998v1 | Sukannya Purkayastha, Anne Lauscher, Iryna Gurevych | 2023-11-07T13:54:01Z | cs.CL | In many domains of argumentation, people's arguments are driven by so-called
attitude roots, i.e., underlying beliefs and world views, and their
corresponding attitude themes. Given the strength of these latent drivers of
arguments, recent work in psychology suggests that instead of directly
countering surface-level reasoning (e.g., falsifying given premises), one
should follow an argumentation style inspired by the Jiu-Jitsu 'soft' combat
system (Hornsey and Fielding, 2017): first, identify an arguer's attitude roots
and themes, and then choose a prototypical rebuttal that is aligned with those
drivers instead of invalidating those. In this work, we are the first to
explore Jiu-Jitsu argumentation for peer review by proposing the novel task of
attitude and theme-guided rebuttal generation. To this end, we enrich an
existing dataset for discourse structure in peer reviews with attitude roots,
attitude themes, and canonical rebuttals. To facilitate this process, we recast
established annotation concepts from the domain of peer reviews (e.g., aspects
a review sentence is relating to) and train domain-specific models. We then
propose strong rebuttal generation strategies, which we benchmark on our novel
dataset for the task of end-to-end attitude and theme-guided rebuttal
generation and two subtasks. | cs.CL | Exploring Jiu-Jitsu Argumentation for Writing Peer Review Rebuttals |
2023-11-07T13:08:55Z | http://arxiv.org/pdf/2311.03969v1 | http://arxiv.org/abs/2311.03969v1 | Gal Ron, Effi Levi, Odelia Oshri, Shaul R. Shenhav | 2023-11-07T13:08:55Z | cs.CL | In this work we propose a novel annotation scheme which factors hate speech
into five separate discursive categories. To evaluate our scheme, we construct
a corpus of over 2.9M Twitter posts containing hateful expressions directed at
Jews, and annotate a sample dataset of 1,050 tweets. We present a statistical
analysis of the annotated dataset as well as discuss annotation examples, and
conclude by discussing promising directions for future work. | cs.CL | Factoring Hate Speech: A New Annotation Framework to Study Hate Speech in Social Media |
2023-11-07T13:03:54Z | http://arxiv.org/pdf/2311.03963v1 | http://arxiv.org/abs/2311.03963v1 | Oseremen O. Uduehi, Razvan C. Bunescu | 2023-11-07T13:03:54Z | cs.CL, cs.AI | We propose a metaphor detection architecture that is structured around two
main modules: an expectation component that estimates representations of
literal word expectations given a context, and a realization component that
computes representations of actual word meanings in context. The overall
architecture is trained to learn expectation-realization (ER) patterns that
characterize metaphorical uses of words. When evaluated on three metaphor
datasets for within distribution, out of distribution, and novel metaphor
generalization, the proposed method is shown to obtain results that are
competitive or better than state-of-the art. Further increases in metaphor
detection accuracy are obtained through ensembling of ER models. | cs.CL | An Expectation-Realization Model for Metaphor Detection |
2023-11-07T12:50:11Z | http://arxiv.org/pdf/2311.03952v2 | http://arxiv.org/abs/2311.03952v2 | Matthew Galbraith | 2024-02-07T22:10:35Z | cs.CL, cs.HC, cs.RO | Spoken dialogue systems have transformed human-machine interaction by
providing real-time responses to queries. However, misunderstandings between
the user and system persist. This study explores the significance of
interactional language in dialogue repair between virtual assistants and users
by analyzing interactions with Google Assistant and Siri, focusing on their
utilization and response to the other-initiated repair strategy "huh?"
prevalent in human-human interaction. Findings reveal several
assistant-generated strategies but an inability to replicate human-like repair
strategies such as "huh?". English and Spanish user acceptability surveys show
differences in users' repair strategy preferences and assistant usage, with
both similarities and disparities among the two surveyed languages. These
results shed light on inequalities between interactional language in
human-human interaction and human-machine interaction, underscoring the need
for further research on the impact of interactional language in human-machine
interaction in English and beyond. | cs.CL | An Analysis of Dialogue Repair in Voice Assistants |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.