publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-02-12T08:45:08Z | http://arxiv.org/pdf/2402.07483v2 | http://arxiv.org/abs/2402.07483v2 | Masoomali Fatehkia, Ji Kim Lucas, Sanjay Chawla | 2024-06-06T14:42:47Z | cs.AI, cs.CL | Large Language Models (LLM) have shown remarkable language capabilities
fueling attempts to integrate them into applications across a wide range of
domains. An important application area is question answering over private
enterprise documents where the main considerations are data security, which
necessitates applications that can be deployed on-prem, limited computational
resources and the need for a robust application that correctly responds to
queries. Retrieval-Augmented Generation (RAG) has emerged as the most prominent
framework for building LLM-based applications. While building a RAG is
relatively straightforward, making it robust and a reliable application
requires extensive customization and relatively deep knowledge of the
application domain. We share our experiences building and deploying an LLM
application for question answering over private organizational documents. Our
application combines the use of RAG with a finetuned open-source LLM.
Additionally, our system, which we call Tree-RAG (T-RAG), uses a tree structure
to represent entity hierarchies within the organization. This is used to
generate a textual description to augment the context when responding to user
queries pertaining to entities within the organization's hierarchy. Our
evaluations, including a Needle in a Haystack test, show that this combination
performs better than a simple RAG or finetuning implementation. Finally, we
share some lessons learned based on our experiences building an LLM application
for real-world use. | cs.AI | T-RAG: Lessons from the LLM Trenches |
2024-02-12T08:14:03Z | http://arxiv.org/pdf/2402.07470v2 | http://arxiv.org/abs/2402.07470v2 | Yazhou Zhang, Mengyao Wang, Chenyu Ren, Qiuchi Li, Prayag Tiwari, Benyou Wang, Jing Qin | 2024-02-16T15:59:27Z | cs.CL | The value of text classification's future research has encountered challenges
and uncertainties, due to the extraordinary efficacy demonstrated by large
language models (LLMs) across numerous downstream NLP tasks. In this era of
open-ended language modeling, where task boundaries are gradually fading, an
urgent question emerges: have we made significant advances in text
classification under the full benefit of LLMs? To answer this question, we
propose RGPT, an adaptive boosting framework tailored to produce a specialized
text classification LLM by recurrently ensembling a pool of strong base
learners. The base learners are constructed by adaptively adjusting the
distribution of training samples and iteratively fine-tuning LLMs with them.
Such base learners are then ensembled to be a specialized text classification
LLM, by recurrently incorporating the historical predictions from the previous
learners. Through a comprehensive empirical comparison, we show that RGPT
significantly outperforms 8 SOTA PLMs and 7 SOTA LLMs on four benchmarks by
1.36% on average. Further evaluation experiments show a clear surpassing of
RGPT over human classification. | cs.CL | Pushing The Limit of LLM Capacity for Text Classification |
2024-02-12T07:11:13Z | http://arxiv.org/pdf/2402.07448v1 | http://arxiv.org/abs/2402.07448v1 | Ahmed Heakl, Youssef Mohamed, Ahmed B. Zaky | 2024-02-12T07:11:13Z | cs.CL, cs.AI, cs.DB, cs.IR, cs.LG | This study presents AraSpider, the first Arabic version of the Spider
dataset, aimed at improving natural language processing (NLP) in the
Arabic-speaking community. Four multilingual translation models were tested for
their effectiveness in translating English to Arabic. Additionally, two models
were assessed for their ability to generate SQL queries from Arabic text. The
results showed that using back translation significantly improved the
performance of both ChatGPT 3.5 and SQLCoder models, which are considered top
performers on the Spider dataset. Notably, ChatGPT 3.5 demonstrated
high-quality translation, while SQLCoder excelled in text-to-SQL tasks. The
study underscores the importance of incorporating contextual schema and
employing back translation strategies to enhance model performance in Arabic
NLP tasks. Moreover, the provision of detailed methodologies for
reproducibility and translation of the dataset into other languages highlights
the research's commitment to promoting transparency and collaborative knowledge
sharing in the field. Overall, these contributions advance NLP research,
empower Arabic-speaking researchers, and enrich the global discourse on
language comprehension and database interrogation. | cs.CL | AraSpider: Democratizing Arabic-to-SQL |
2024-02-12T07:03:14Z | http://arxiv.org/pdf/2402.07446v3 | http://arxiv.org/abs/2402.07446v3 | Surangika Ranathunga, Nisansa de Silva, Menan Velayuthan, Aloka Fernando, Charitha Rathnayake | 2024-06-14T08:50:53Z | cs.CL | We conducted a detailed analysis on the quality of web-mined corpora for two
low-resource languages (making three language pairs, English-Sinhala,
English-Tamil and Sinhala-Tamil). We ranked each corpus according to a
similarity measure and carried out an intrinsic and extrinsic evaluation on
different portions of this ranked corpus. We show that there are significant
quality differences between different portions of web-mined corpora and that
the quality varies across languages and datasets. We also show that, for some
web-mined datasets, Neural Machine Translation (NMT) models trained with their
highest-ranked 25k portion can be on par with human-curated datasets. | cs.CL | Quality Does Matter: A Detailed Look at the Quality and Utility of Web-Mined Parallel Corpora |
2024-02-12T06:21:35Z | http://arxiv.org/pdf/2402.07432v1 | http://arxiv.org/abs/2402.07432v1 | Guanyi Chen, Fahime Same, Kees van Deemter | 2024-02-12T06:21:35Z | cs.CL | Recently, a human evaluation study of Referring Expression Generation (REG)
models had an unexpected conclusion: on \textsc{webnlg}, Referring Expressions
(REs) generated by the state-of-the-art neural models were not only
indistinguishable from the REs in \textsc{webnlg} but also from the REs
generated by a simple rule-based system. Here, we argue that this limitation
could stem from the use of a purely ratings-based human evaluation (which is a
common practice in Natural Language Generation). To investigate these issues,
we propose an intrinsic task-based evaluation for REG models, in which, in
addition to rating the quality of REs, participants were asked to accomplish
two meta-level tasks. One of these tasks concerns the referential success of
each RE; the other task asks participants to suggest a better alternative for
each RE. The outcomes suggest that, in comparison to previous evaluations, the
new evaluation protocol assesses the performance of each REG model more
comprehensively and makes the participants' ratings more reliable and
discriminable. | cs.CL | Intrinsic Task-based Evaluation for Referring Expression Generation |
2024-02-12T06:15:24Z | http://arxiv.org/pdf/2402.07431v2 | http://arxiv.org/abs/2402.07431v2 | Ragib Amin Nihal, Tran Dong Huu Quoc, Lin Zirui, Xu Yimimg, Liu Haoran, An Zhaoyi, Kyou Ma | 2024-02-13T06:34:48Z | cs.CL, cs.CY | SALAD is an AI-driven language-learning application designed to help
foreigners learn Japanese. It offers translations in Kanji-Kana-Romaji, speech
recognition, translated audio, vocabulary tracking, grammar explanations, and
songs generated from newly learned words. The app targets beginners and
intermediate learners, aiming to make language acquisition more accessible and
enjoyable. SALAD uses daily translations to enhance fluency and comfort in
communication with native speakers. The primary objectives include effective
Japanese language learning, user engagement, and progress tracking. A survey by
us found that 39% of foreigners in Japan face discomfort in conversations with
Japanese speakers. Over 60% of foreigners expressed confidence in SALAD's
ability to enhance their Japanese language skills. The app uses large language
models, speech recognition, and diffusion models to bridge the language gap and
foster a more inclusive community in Japan. | cs.CL | SALAD: Smart AI Language Assistant Daily |
2024-02-12T04:50:31Z | http://arxiv.org/pdf/2402.07405v1 | http://arxiv.org/abs/2402.07405v1 | Xiao Zhang, Ruoyu Xiang, Chenhan Yuan, Duanyu Feng, Weiguang Han, Alejandro Lopez-Lira, Xiao-Yang Liu, Sophia Ananiadou, Min Peng, Jimin Huang, Qianqian Xie | 2024-02-12T04:50:31Z | cs.CL | Despite Spanish's pivotal role in the global finance industry, a pronounced
gap exists in Spanish financial natural language processing (NLP) and
application studies compared to English, especially in the era of large
language models (LLMs). To bridge this gap, we unveil Tois\'on de Oro, the
first bilingual framework that establishes instruction datasets, finetuned
LLMs, and evaluation benchmark for financial LLMs in Spanish joint with
English. We construct a rigorously curated bilingual instruction dataset
including over 144K Spanish and English samples from 15 datasets covering 7
tasks. Harnessing this, we introduce FinMA-ES, an LLM designed for bilingual
financial applications. We evaluate our model and existing LLMs using FLARE-ES,
the first comprehensive bilingual evaluation benchmark with 21 datasets
covering 9 tasks. The FLARE-ES benchmark results reveal a significant
multilingual performance gap and bias in existing LLMs. FinMA-ES models surpass
SOTA LLMs such as GPT-4 in Spanish financial tasks, due to strategic
instruction tuning and leveraging data from diverse linguistic resources,
highlighting the positive impact of cross-linguistic transfer. All our
datasets, models, and benchmarks have been released. | cs.CL | Dólares or Dollars? Unraveling the Bilingual Prowess of Financial LLMs Between Spanish and English |
2024-02-12T04:32:33Z | http://arxiv.org/pdf/2402.07401v1 | http://arxiv.org/abs/2402.07401v1 | Kyungha Kim, Sangyun Lee, Kung-Hsiang Huang, Hou Pong Chan, Manling Li, Heng Ji | 2024-02-12T04:32:33Z | cs.CL | Fact-checking research has extensively explored verification but less so the
generation of natural-language explanations, crucial for user trust. While
Large Language Models (LLMs) excel in text generation, their capability for
producing faithful explanations in fact-checking remains underexamined. Our
study investigates LLMs' ability to generate such explanations, finding that
zero-shot prompts often result in unfaithfulness. To address these challenges,
we propose the Multi-Agent Debate Refinement (MADR) framework, leveraging
multiple LLMs as agents with diverse roles in an iterative refining process
aimed at enhancing faithfulness in generated explanations. MADR ensures that
the final explanation undergoes rigorous validation, significantly reducing the
likelihood of unfaithful elements and aligning closely with the provided
evidence. Experimental results demonstrate that MADR significantly improves the
faithfulness of LLM-generated explanations to the evidence, advancing the
credibility and trustworthiness of these explanations. | cs.CL | Can LLMs Produce Faithful Explanations For Fact-checking? Towards Faithful Explainable Fact-Checking via Multi-Agent Debate |
2024-02-12T04:10:09Z | http://arxiv.org/pdf/2402.07397v2 | http://arxiv.org/abs/2402.07397v2 | George Boateng | 2024-04-26T23:31:04Z | cs.CY, cs.CL, cs.HC | Across the African continent, students grapple with various educational
challenges, including limited access to essential resources such as computers,
internet connectivity, reliable electricity, and a shortage of qualified
teachers. Despite these challenges, recent advances in AI such as BERT, and
GPT-4 have demonstrated their potential for advancing education. Yet, these AI
tools tend to be deployed and evaluated predominantly within the context of
Western educational settings, with limited attention directed towards the
unique needs and challenges faced by students in Africa. In this chapter, we
discuss challenges with using AI to advance education across Africa. Then, we
describe our work developing and deploying AI in Education tools in Africa for
science and computing education: (1) SuaCode, an AI-powered app that enables
Africans to learn to code using their smartphones, (2) AutoGrad, an automated
grading, and feedback tool for graphical and interactive coding assignments,
(3) a tool for code plagiarism detection that shows visual evidence of
plagiarism, (4) Kwame, a bilingual AI teaching assistant for coding courses,
(5) Kwame for Science, a web-based AI teaching assistant that provides instant
answers to students' science questions and (6) Brilla AI, an AI contestant for
the National Science and Maths Quiz competition. Finally, we discuss potential
opportunities to leverage AI to advance education across Africa. | cs.CY | Leveraging AI to Advance Science and Computing Education across Africa: Challenges, Progress and Opportunities |
2024-02-12T03:05:54Z | http://arxiv.org/pdf/2402.07386v1 | http://arxiv.org/abs/2402.07386v1 | Qingkai Zeng, Yuyang Bai, Zhaoxuan Tan, Shangbin Feng, Zhenwen Liang, Zhihan Zhang, Meng Jiang | 2024-02-12T03:05:54Z | cs.CL | Automatic taxonomy induction is crucial for web search, recommendation
systems, and question answering. Manual curation of taxonomies is expensive in
terms of human effort, making automatic taxonomy construction highly desirable.
In this work, we introduce Chain-of-Layer which is an in-context learning
framework designed to induct taxonomies from a given set of entities.
Chain-of-Layer breaks down the task into selecting relevant candidate entities
in each layer and gradually building the taxonomy from top to bottom. To
minimize errors, we introduce the Ensemble-based Ranking Filter to reduce the
hallucinated content generated at each iteration. Through extensive
experiments, we demonstrate that Chain-of-Layer achieves state-of-the-art
performance on four real-world benchmarks. | cs.CL | Chain-of-Layer: Iteratively Prompting Large Language Models for Taxonomy Induction from Limited Examples |
2024-02-12T02:58:10Z | http://arxiv.org/pdf/2402.07383v2 | http://arxiv.org/abs/2402.07383v2 | Naoyuki Kanda, Xiaofei Wang, Sefik Emre Eskimez, Manthan Thakker, Hemin Yang, Zirun Zhu, Min Tang, Canrun Li, Chung-Hsien Tsai, Zhen Xiao, Yufei Xia, Jinzhu Li, Yanqing Liu, Sheng Zhao, Michael Zeng | 2024-03-04T19:15:29Z | eess.AS, cs.CL, cs.LG, cs.SD | Laughter is one of the most expressive and natural aspects of human speech,
conveying emotions, social cues, and humor. However, most text-to-speech (TTS)
systems lack the ability to produce realistic and appropriate laughter sounds,
limiting their applications and user experience. While there have been prior
works to generate natural laughter, they fell short in terms of controlling the
timing and variety of the laughter to be generated. In this work, we propose
ELaTE, a zero-shot TTS that can generate natural laughing speech of any speaker
based on a short audio prompt with precise control of laughter timing and
expression. Specifically, ELaTE works on the audio prompt to mimic the voice
characteristic, the text prompt to indicate the contents of the generated
speech, and the input to control the laughter expression, which can be either
the start and end times of laughter, or the additional audio prompt that
contains laughter to be mimicked. We develop our model based on the foundation
of conditional flow-matching-based zero-shot TTS, and fine-tune it with
frame-level representation from a laughter detector as additional conditioning.
With a simple scheme to mix small-scale laughter-conditioned data with
large-scale pre-training data, we demonstrate that a pre-trained zero-shot TTS
model can be readily fine-tuned to generate natural laughter with precise
controllability, without losing any quality of the pre-trained zero-shot TTS
model. Through objective and subjective evaluations, we show that ELaTE can
generate laughing speech with significantly higher quality and controllability
compared to conventional models. See https://aka.ms/elate/ for demo samples. | eess.AS | Making Flow-Matching-Based Zero-Shot Text-to-Speech Laugh as You Like |
2024-02-12T01:55:51Z | http://arxiv.org/pdf/2402.07368v1 | http://arxiv.org/abs/2402.07368v1 | Gabriel Simmons, Vladislav Savinov | 2024-02-12T01:55:51Z | cs.LG, cs.AI, cs.CL, cs.CY | This study evaluates the ability of Large Language Model (LLM)-based
Subpopulation Representative Models (SRMs) to generalize from empirical data,
utilizing in-context learning with data from the 2016 and 2020 American
National Election Studies. We explore generalization across response variables
and demographic subgroups. While conditioning with empirical data improves
performance on the whole, the benefit of in-context learning varies
considerably across demographics, sometimes hurting performance for one
demographic while helping performance for others. The inequitable benefits of
in-context learning for SRM present a challenge for practitioners implementing
SRMs, and for decision-makers who might come to rely on them. Our work
highlights a need for fine-grained benchmarks captured from diverse
subpopulations that test not only fidelity but generalization. | cs.LG | Assessing Generalization for Subpopulation Representative Modeling via In-Context Learning |
2024-02-11T22:58:49Z | http://arxiv.org/pdf/2402.07321v1 | http://arxiv.org/abs/2402.07321v1 | Bilal Chughtai, Alan Cooney, Neel Nanda | 2024-02-11T22:58:49Z | cs.LG, cs.CL | How do transformer-based large language models (LLMs) store and retrieve
knowledge? We focus on the most basic form of this task -- factual recall,
where the model is tasked with explicitly surfacing stored facts in prompts of
form `Fact: The Colosseum is in the country of'. We find that the mechanistic
story behind factual recall is more complex than previously thought. It
comprises several distinct, independent, and qualitatively different mechanisms
that additively combine, constructively interfering on the correct attribute.
We term this generic phenomena the additive motif: models compute through
summing up multiple independent contributions. Each mechanism's contribution
may be insufficient alone, but summing results in constructive interfere on the
correct answer. In addition, we extend the method of direct logit attribution
to attribute an attention head's output to individual source tokens. We use
this technique to unpack what we call `mixed heads' -- which are themselves a
pair of two separate additive updates from different source tokens. | cs.LG | Summing Up the Facts: Additive Mechanisms Behind Factual Recall in LLMs |
2024-02-11T22:40:12Z | http://arxiv.org/pdf/2402.07319v1 | http://arxiv.org/abs/2402.07319v1 | Lichang Chen, Chen Zhu, Davit Soselia, Jiuhai Chen, Tianyi Zhou, Tom Goldstein, Heng Huang, Mohammad Shoeybi, Bryan Catanzaro | 2024-02-11T22:40:12Z | cs.LG, cs.AI, cs.CL | In this work, we study the issue of reward hacking on the response length, a
challenge emerging in Reinforcement Learning from Human Feedback (RLHF) on
LLMs. A well-formatted, verbose but less helpful response from the LLMs can
often deceive LLMs or even human evaluators to achieve high scores. The same
issue also holds for some reward models in RL. To address the challenges in
both training and evaluation, we establish a more reliable evaluation protocol
for comparing different training configurations, which inspects the trade-off
between LLM evaluation score and response length obtained by varying training
hyperparameters. Based on this evaluation, we conduct large-scale studies,
where the results shed insights into the efficacy of hyperparameters and tricks
used in RL on mitigating length bias. We further propose to improve the reward
model by jointly training two linear heads on shared feature representations to
predict the rewards, one trained to correlate with length, and the other
trained to decorrelate with length and therefore focus more on the actual
content. We then discard the length head in RL to prevent reward hacking on
length. Experiments demonstrate that our approach almost eliminates the reward
correlation with length, and improves the obtained policy by a significant
margin. | cs.LG | ODIN: Disentangled Reward Mitigates Hacking in RLHF |
2024-02-11T21:16:26Z | http://arxiv.org/pdf/2402.07309v3 | http://arxiv.org/abs/2402.07309v3 | Adrián Bazaga, Pietro Liò, Gos Micklem | 2024-05-25T12:10:38Z | cs.LG, cs.CL, stat.ML | Hypergraphs are characterized by complex topological structure, representing
higher-order interactions among multiple entities through hyperedges. Lately,
hypergraph-based deep learning methods to learn informative data
representations for the problem of node classification on text-attributed
hypergraphs have garnered increasing research attention. However, existing
methods struggle to simultaneously capture the full extent of hypergraph
structural information and the rich linguistic attributes inherent in the nodes
attributes, which largely hampers their effectiveness and generalizability. To
overcome these challenges, we explore ways to further augment a pretrained BERT
model with specialized hypergraph-aware layers for the task of node
classification. Such layers introduce higher-order structural inductive bias
into the language model, thus improving the model's capacity to harness both
higher-order context information from the hypergraph structure and semantic
information present in text. In this paper, we propose a new architecture,
HyperBERT, a mixed text-hypergraph model which simultaneously models hypergraph
relational structure while maintaining the high-quality text encoding
capabilities of a pre-trained BERT. Notably, HyperBERT presents results that
achieve a new state-of-the-art on five challenging text-attributed hypergraph
node classification benchmarks. | cs.LG | HyperBERT: Mixing Hypergraph-Aware Layers with Language Models for Node Classification on Text-Attributed Hypergraphs |
2024-02-11T19:14:28Z | http://arxiv.org/pdf/2402.07283v1 | http://arxiv.org/abs/2402.07283v1 | Chao Wang, Zhuo Chen, Ziyan Zhang, Chiyi Li, Kai Song | 2024-02-11T19:14:28Z | cs.LG, cs.AI, cs.CL | In this paper, we address the challenge of learning with limited fault data
for power transformers. Traditional operation and maintenance tools lack
effective predictive capabilities for potential faults. The scarcity of
extensive fault data makes it difficult to apply machine learning techniques
effectively. To solve this problem, we propose a novel approach that leverages
the knowledge graph (KG) technology in combination with gradient boosting
decision trees (GBDT). This method is designed to efficiently learn from a
small set of high-dimensional data, integrating various factors influencing
transformer faults and historical operational data. Our approach enables
accurate safe state assessments and fault analyses of power transformers
despite the limited fault characteristic data. Experimental results demonstrate
that this method outperforms other learning approaches in prediction accuracy,
such as artificial neural networks (ANN) and logistic regression (LR).
Furthermore, it offers significant improvements in progressiveness,
practicality, and potential for widespread application. | cs.LG | Power Transformer Fault Prediction Based on Knowledge Graphs |
2024-02-11T19:13:26Z | http://arxiv.org/pdf/2402.07282v2 | http://arxiv.org/abs/2402.07282v2 | Ryan Liu, Theodore R. Sumers, Ishita Dasgupta, Thomas L. Griffiths | 2024-02-13T14:21:02Z | cs.CL, cs.AI, cs.LG | In day-to-day communication, people often approximate the truth - for
example, rounding the time or omitting details - in order to be maximally
helpful to the listener. How do large language models (LLMs) handle such
nuanced trade-offs? To address this question, we use psychological models and
experiments designed to characterize human behavior to analyze LLMs. We test a
range of LLMs and explore how optimization for human preferences or
inference-time reasoning affects these trade-offs. We find that reinforcement
learning from human feedback improves both honesty and helpfulness, while
chain-of-thought prompting skews LLMs towards helpfulness over honesty.
Finally, GPT-4 Turbo demonstrates human-like response patterns including
sensitivity to the conversational framing and listener's decision context. Our
findings reveal the conversational values internalized by LLMs and suggest that
even these abstract values can, to a degree, be steered by zero-shot prompting. | cs.CL | How do Large Language Models Navigate Conflicts between Honesty and Helpfulness? |
2024-02-11T18:27:14Z | http://arxiv.org/pdf/2402.07271v1 | http://arxiv.org/abs/2402.07271v1 | Jiangnan Li, Qiujing Wang, Liyan Xu, Wenjie Pang, Mo Yu, Zheng Lin, Weiping Wang, Jie Zhou | 2024-02-11T18:27:14Z | cs.CL | Similar to the "previously-on" scenes in TV shows, recaps can help book
reading by recalling the readers' memory about the important elements in
previous texts to better understand the ongoing plot. Despite its usefulness,
this application has not been well studied in the NLP community. We propose the
first benchmark on this useful task called Recap Snippet Identification with a
hand-crafted evaluation dataset. Our experiments show that the proposed task is
challenging to PLMs, LLMs, and proposed methods as the task requires a deep
understanding of the plot correlation between snippets. | cs.CL | Previously on the Stories: Recap Snippet Identification for Story Reading |
2024-02-11T18:26:18Z | http://arxiv.org/pdf/2402.07270v2 | http://arxiv.org/abs/2402.07270v2 | Simon Ging, María A. Bravo, Thomas Brox | 2024-05-05T20:34:28Z | cs.CV, cs.CL, cs.LG | The evaluation of text-generative vision-language models is a challenging yet
crucial endeavor. By addressing the limitations of existing Visual Question
Answering (VQA) benchmarks and proposing innovative evaluation methodologies,
our research seeks to advance our understanding of these models' capabilities.
We propose a novel VQA benchmark based on well-known visual classification
datasets which allows a granular evaluation of text-generative vision-language
models and their comparison with discriminative vision-language models. To
improve the assessment of coarse answers on fine-grained classification tasks,
we suggest using the semantic hierarchy of the label space to ask automatically
generated follow-up questions about the ground-truth category. Finally, we
compare traditional NLP and LLM-based metrics for the problem of evaluating
model predictions given ground-truth answers. We perform a human evaluation
study upon which we base our decision on the final metric. We apply our
benchmark to a suite of vision-language models and show a detailed comparison
of their abilities on object, action, and attribute classification. Our
contributions aim to lay the foundation for more precise and meaningful
assessments, facilitating targeted progress in the exciting field of
vision-language modeling. | cs.CV | Open-ended VQA benchmarking of Vision-Language models by exploiting Classification datasets and their semantic hierarchy |
2024-02-11T18:09:50Z | http://arxiv.org/pdf/2402.07262v1 | http://arxiv.org/abs/2402.07262v1 | Mithun Das, Saurabh Kumar Pandey, Shivansh Sethi, Punyajoy Saha, Animesh Mukherjee | 2024-02-11T18:09:50Z | cs.CL, cs.HC | With the rise of online abuse, the NLP community has begun investigating the
use of neural architectures to generate counterspeech that can "counter" the
vicious tone of such abusive speech and dilute/ameliorate their rippling effect
over the social network. However, most of the efforts so far have been
primarily focused on English. To bridge the gap for low-resource languages such
as Bengali and Hindi, we create a benchmark dataset of 5,062 abusive
speech/counterspeech pairs, of which 2,460 pairs are in Bengali and 2,602 pairs
are in Hindi. We implement several baseline models considering various
interlingual transfer mechanisms with different configurations to generate
suitable counterspeech to set up an effective benchmark. We observe that the
monolingual setup yields the best performance. Further, using synthetic
transfer, language models can generate counterspeech to some extent;
specifically, we notice that transferability is better when languages belong to
the same language family. | cs.CL | Low-Resource Counterspeech Generation for Indic Languages: The Case of Bengali and Hindi |
2024-02-11T17:46:33Z | http://arxiv.org/pdf/2402.07255v1 | http://arxiv.org/abs/2402.07255v1 | Parsheeta Roy, Ji-Eun Han, Srishti Chouhan, Bhaavanaa Thumu | 2024-02-11T17:46:33Z | cs.CL, cs.CV | Sign language to text is a crucial technology that can break down
communication barriers for individuals with hearing difficulties. We replicate
and try to improve on a recently published study. We evaluate models using BLEU
and rBLEU metrics to ensure translation quality. During our ablation study, we
found that the model's performance is significantly influenced by optimizers,
activation functions, and label smoothing. Further research aims to refine
visual feature capturing, enhance decoder utilization, and integrate
pre-trained decoders for better translation outcomes. Our source code is
available to facilitate replication of our results and encourage future
research. | cs.CL | American Sign Language Video to Text Translation |
2024-02-11T15:50:35Z | http://arxiv.org/pdf/2402.07233v1 | http://arxiv.org/abs/2402.07233v1 | Peng Wang, Xiang Wei, Fangxu Hu, Wenjuan Han | 2024-02-11T15:50:35Z | cs.CL, cs.AI | Natural language processing (NLP) is a key component of intelligent
transportation systems (ITS), but it faces many challenges in the
transportation domain, such as domain-specific knowledge and data, and
multi-modal inputs and outputs. This paper presents TransGPT, a novel
(multi-modal) large language model for the transportation domain, which
consists of two independent variants: TransGPT-SM for single-modal data and
TransGPT-MM for multi-modal data. TransGPT-SM is finetuned on a single-modal
Transportation dataset (STD) that contains textual data from various sources in
the transportation domain. TransGPT-MM is finetuned on a multi-modal
Transportation dataset (MTD) that we manually collected from three areas of the
transportation domain: driving tests, traffic signs, and landmarks. We evaluate
TransGPT on several benchmark datasets for different tasks in the
transportation domain, and show that it outperforms baseline models on most
tasks. We also showcase the potential applications of TransGPT for traffic
analysis and modeling, such as generating synthetic traffic scenarios,
explaining traffic phenomena, answering traffic-related questions, providing
traffic recommendations, and generating traffic reports. This work advances the
state-of-the-art of NLP in the transportation domain and provides a useful tool
for ITS researchers and practitioners. | cs.CL | TransGPT: Multi-modal Generative Pre-trained Transformer for Transportation |
2024-02-11T14:21:37Z | http://arxiv.org/pdf/2402.07214v3 | http://arxiv.org/abs/2402.07214v3 | Shanshan Xu, T. Y. S. S Santosh, Oana Ichim, Barbara Plank, Matthias Grabmair | 2024-06-06T10:53:50Z | cs.CL | In legal decisions, split votes (SV) occur when judges cannot reach a
unanimous decision, posing a difficulty for lawyers who must navigate diverse
legal arguments and opinions. In high-stakes domains, understanding the
alignment of perceived difficulty between humans and AI systems is crucial to
build trust. However, existing NLP calibration methods focus on a classifier's
awareness of predictive performance, measured against the human majority class,
overlooking inherent human label variation (HLV). This paper explores split
votes as naturally observable human disagreement and value pluralism. We
collect judges' vote distributions from the European Court of Human Rights
(ECHR), and present SV-ECHR, a case outcome classification (COC) dataset with
SV information. We build a taxonomy of disagreement with SV-specific
subcategories. We further assess the alignment of perceived difficulty between
models and humans, as well as confidence- and human-calibration of COC models.
We observe limited alignment with the judge vote distribution. To our
knowledge, this is the first systematic exploration of calibration to human
judgements in legal NLP. Our study underscores the necessity for further
research on measuring and enhancing model calibration considering HLV in legal
decision tasks. | cs.CL | Through the Lens of Split Vote: Exploring Disagreement, Difficulty and Calibration in Legal Case Outcome Classification |
2024-02-11T13:41:17Z | http://arxiv.org/pdf/2402.15518v1 | http://arxiv.org/abs/2402.15518v1 | Gonzalo Martínez, José Alberto Hernández, Javier Conde, Pedro Reviriego, Elena Merino | 2024-02-11T13:41:17Z | cs.CL | The performance of conversational Large Language Models (LLMs) in general,
and of ChatGPT in particular, is currently being evaluated on many different
tasks, from logical reasoning or maths to answering questions on a myriad of
topics. Instead, much less attention is being devoted to the study of the
linguistic features of the texts generated by these LLMs. This is surprising
since LLMs are models for language, and understanding how they use the language
is important. Indeed, conversational LLMs are poised to have a significant
impact on the evolution of languages as they may eventually dominate the
creation of new text. This means that for example, if conversational LLMs do
not use a word it may become less and less frequent and eventually stop being
used altogether. Therefore, evaluating the linguistic features of the text they
produce and how those depend on the model parameters is the first step toward
understanding the potential impact of conversational LLMs on the evolution of
languages. In this paper, we consider the evaluation of the lexical richness of
the text generated by LLMs and how it depends on the model parameters. A
methodology is presented and used to conduct a comprehensive evaluation of
lexical richness using ChatGPT as a case study. The results show how lexical
richness depends on the version of ChatGPT and some of its parameters, such as
the presence penalty, or on the role assigned to the model. The dataset and
tools used in our analysis are released under open licenses with the goal of
drawing the much-needed attention to the evaluation of the linguistic features
of LLM-generated text. | cs.CL | Beware of Words: Evaluating the Lexical Richness of Conversational Large Language Models |
2024-02-11T13:30:53Z | http://arxiv.org/pdf/2402.07204v2 | http://arxiv.org/abs/2402.07204v2 | Yihong Tang, Zhaokai Wang, Ao Qu, Yihao Yan, Kebing Hou, Dingyi Zhuang, Xiaotong Guo, Jinhua Zhao, Zhan Zhao, Wei Ma | 2024-05-23T10:24:00Z | cs.AI, cs.CL, cs.LG | In this paper, we introduce the novel task of Open-domain Urban Itinerary
Planning (OUIP), a paradigm designed to generate personalized urban itineraries
from user requests articulated in natural language. This approach is different
from traditional itinerary planning, which often restricts the granularity of
user inputs, thus hindering genuine personalization. To this end, we present
ItiNera, an OUIP system that synergizes spatial optimization with large
language models (LLMs) to provide services that customize urban itineraries
based on users' needs. Upon receiving the user's itinerary request, the LLM
first decomposes it into detailed components, identifying key requirements,
including preferences and dislikes. Then, we use these specifics to select
candidate POIs from a large-scale collection using embedding-based
Preference-aware POI Retrieval. Finally, a preference score-based Cluster-aware
Spatial Optimization module clusters, filters, and orders these POIs, followed
by the LLM for detailed POI selection and organization to craft a personalized,
spatially coherent itinerary. Moreover, we created an LLM-based pipeline to
update and personalize a user-owned POI database. This ensures up-to-date POI
information, supports itinerary planning, pre-trip research, POI collection,
recommendations, and more. To the best of our knowledge, this study marks the
first integration of LLMs to innovate itinerary planning, with potential
extensions for various urban travel and exploration activities. Offline and
online evaluations demonstrate the capacity of our system to deliver more
responsive, personalized, and spatially coherent itineraries than current
solutions. Our system, deployed on an online platform, has attracted thousands
of users for their urban travel planning. | cs.AI | Synergizing Spatial Optimization with Large Language Models for Open-Domain Urban Itinerary Planning |
2024-02-11T12:25:41Z | http://arxiv.org/pdf/2402.07179v2 | http://arxiv.org/abs/2402.07179v2 | Zhibo Hu, Chen Wang, Yanfeng Shu, Helen, Paik, Liming Zhu | 2024-06-20T14:07:27Z | cs.CL, cs.IR, I.2.7; H.3.3 | The robustness of large language models (LLMs) becomes increasingly important
as their use rapidly grows in a wide range of domains. Retrieval-Augmented
Generation (RAG) is considered as a means to improve the trustworthiness of
text generation from LLMs. However, how the outputs from RAG-based LLMs are
affected by slightly different inputs is not well studied. In this work, we
find that the insertion of even a short prefix to the prompt leads to the
generation of outputs far away from factually correct answers. We
systematically evaluate the effect of such prefixes on RAG by introducing a
novel optimization technique called Gradient Guided Prompt Perturbation (GGPP).
GGPP achieves a high success rate in steering outputs of RAG-based LLMs to
targeted wrong answers. It can also cope with instructions in the prompts
requesting to ignore irrelevant context. We also exploit LLMs' neuron
activation difference between prompts with and without GGPP perturbations to
give a method that improves the robustness of RAG-based LLMs through a highly
effective detector trained on neuron activation triggered by GGPP generated
prompts. Our evaluation on open-sourced LLMs demonstrates the effectiveness of
our methods. | cs.CL | Prompt Perturbation in Retrieval-Augmented Generation based Large Language Models |
2024-02-11T11:03:04Z | http://arxiv.org/pdf/2402.07157v2 | http://arxiv.org/abs/2402.07157v2 | Xidong Feng, Ziyu Wan, Mengyue Yang, Ziyan Wang, Girish A. Koushik, Yali Du, Ying Wen, Jun Wang | 2024-02-14T19:59:05Z | cs.CL, cs.AI, cs.LG | Reinforcement Learning (RL) has shown remarkable abilities in learning
policies for decision-making tasks. However, RL is often hindered by issues
such as low sample efficiency, lack of interpretability, and sparse supervision
signals. To tackle these limitations, we take inspiration from the human
learning process and introduce Natural Language Reinforcement Learning (NLRL),
which innovatively combines RL principles with natural language representation.
Specifically, NLRL redefines RL concepts like task objectives, policy, value
function, Bellman equation, and policy iteration in natural language space. We
present how NLRL can be practically implemented with the latest advancements in
large language models (LLMs) like GPT-4. Initial experiments over tabular MDPs
demonstrate the effectiveness, efficiency, and also interpretability of the
NLRL framework. | cs.CL | Natural Language Reinforcement Learning |
2024-02-11T10:23:34Z | http://arxiv.org/pdf/2402.07148v2 | http://arxiv.org/abs/2402.07148v2 | Eric L. Buehler, Markus J. Buehler | 2024-03-30T20:18:54Z | cond-mat.soft, cond-mat.dis-nn, cs.AI, cs.CL, cs.LG, q-bio.QM | We report a mixture of expert strategy to create fine-tuned large language
models using a deep layer-wise token-level approach based on low-rank
adaptation (LoRA). Starting with a set of pre-trained LoRA adapters, our gating
strategy uses the hidden states to dynamically mix adapted layers, allowing the
resulting X-LoRA model to draw upon different capabilities and create
never-before-used deep layer-wise combinations to solve tasks. The design is
inspired by the biological principles of universality and diversity, where
neural network building blocks are reused in different hierarchical
manifestations. Hence, the X-LoRA model can be easily implemented for any
existing large language model (LLM) without a need for modifications of the
underlying structure. We develop a tailored X-LoRA model that offers scientific
capabilities including forward/inverse analysis tasks and enhanced reasoning
capability, focused on biomaterial analysis, protein mechanics and design. The
impact of this work include access to readily expandable and adaptable models
with strong domain knowledge and the capability to integrate across areas of
knowledge. Featuring experts in biology, mathematics, reasoning, bio-inspired
materials, mechanics and materials, chemistry, protein biophysics, mechanics
and quantum-mechanics based molecular properties, we conduct a series of
physics-focused case studies. We examine knowledge recall, protein mechanics
forward/inverse tasks, protein design, adversarial agentic modeling including
ontological knowledge graph construction, as well as molecular design. The
model is capable not only of making quantitative predictions of nanomechanical
properties of proteins or quantum mechanical molecular properties, but also
reasons over the results and correctly predicts likely mechanisms that explain
distinct molecular behaviors. | cond-mat.soft | X-LoRA: Mixture of Low-Rank Adapter Experts, a Flexible Framework for Large Language Models with Applications in Protein Mechanics and Molecular Design |
2024-02-11T03:27:22Z | http://arxiv.org/pdf/2402.07092v3 | http://arxiv.org/abs/2402.07092v3 | Haonan Chen, Zhicheng Dou, Kelong Mao, Jiongnan Liu, Ziliang Zhao | 2024-06-04T02:09:15Z | cs.CL, cs.IR | Conversational search utilizes muli-turn natural language contexts to
retrieve relevant passages. Existing conversational dense retrieval models
mostly view a conversation as a fixed sequence of questions and responses,
overlooking the severe data sparsity problem -- that is, users can perform a
conversation in various ways, and these alternate conversations are unrecorded.
Consequently, they often struggle to generalize to diverse conversations in
real-world scenarios. In this work, we propose a framework for generalizing
Conversational dense retrieval via LLM-cognition data Augmentation (ConvAug).
ConvAug first generates multi-level augmented conversations to capture the
diverse nature of conversational contexts. Inspired by human cognition, we
devise a cognition-aware process to mitigate the generation of false positives,
false negatives, and hallucinations. Moreover, we develop a difficulty-adaptive
sample filter that selects challenging samples for complex conversations,
thereby giving the model a larger learning space. A contrastive learning
objective is then employed to train a better conversational context encoder.
Extensive experiments conducted on four public datasets, under both normal and
zero-shot settings, demonstrate the effectiveness, generalizability, and
applicability of ConvAug. The code is released at
https://github.com/haon-chen/ConvAug. | cs.CL | Generalizing Conversational Dense Retrieval via LLM-Cognition Data Augmentation |
2024-02-11T02:26:43Z | http://arxiv.org/pdf/2402.07085v1 | http://arxiv.org/abs/2402.07085v1 | Kenichi Fujita, Atsushi Ando, Yusuke Ijima | 2024-02-11T02:26:43Z | cs.SD, cs.CL, cs.LG, eess.AS | This paper proposes a speech rhythm-based method for speaker embeddings to
model phoneme duration using a few utterances by the target speaker. Speech
rhythm is one of the essential factors among speaker characteristics, along
with acoustic features such as F0, for reproducing individual utterances in
speech synthesis. A novel feature of the proposed method is the rhythm-based
embeddings extracted from phonemes and their durations, which are known to be
related to speaking rhythm. They are extracted with a speaker identification
model similar to the conventional spectral feature-based one. We conducted
three experiments, speaker embeddings generation, speech synthesis with
generated embeddings, and embedding space analysis, to evaluate the
performance. The proposed method demonstrated a moderate speaker identification
performance (15.2% EER), even with only phonemes and their duration
information. The objective and subjective evaluation results demonstrated that
the proposed method can synthesize speech with speech rhythm closer to the
target speaker than the conventional method. We also visualized the embeddings
to evaluate the relationship between the distance of the embeddings and the
perceptual similarity. The visualization of the embedding space and the
relation analysis between the closeness indicated that the distribution of
embeddings reflects the subjective and objective similarity. | cs.SD | Speech Rhythm-Based Speaker Embeddings Extraction from Phonemes and Phoneme Duration for Multi-Speaker Speech Synthesis |
2024-02-11T01:37:48Z | http://arxiv.org/pdf/2402.07081v1 | http://arxiv.org/abs/2402.07081v1 | Nischal Ashok Kumar, Andrew Lan | 2024-02-11T01:37:48Z | cs.CL, cs.SE | In computer science education, test cases are an integral part of programming
assignments since they can be used as assessment items to test students'
programming knowledge and provide personalized feedback on student-written
code. The goal of our work is to propose a fully automated approach for test
case generation that can accurately measure student knowledge, which is
important for two reasons. First, manually constructing test cases requires
expert knowledge and is a labor-intensive process. Second, developing test
cases for students, especially those who are novice programmers, is
significantly different from those oriented toward professional-level software
developers. Therefore, we need an automated process for test case generation to
assess student knowledge and provide feedback. In this work, we propose a large
language model-based approach to automatically generate test cases and show
that they are good measures of student knowledge, using a publicly available
dataset that contains student-written Java code. We also discuss future
research directions centered on using test cases to help students. | cs.CL | Using Large Language Models for Student-Code Guided Test Case Generation in Computer Science Education |
2024-02-11T00:00:05Z | http://arxiv.org/pdf/2402.07069v1 | http://arxiv.org/abs/2402.07069v1 | Shayan Meshkat Alsadat, Jean-Raphael Gaglione, Daniel Neider, Ufuk Topcu, Zhe Xu | 2024-02-11T00:00:05Z | cs.LG, cs.AI, cs.CL | We present LARL-RM (Large language model-generated Automaton for
Reinforcement Learning with Reward Machine) algorithm in order to encode
high-level knowledge into reinforcement learning using automaton to expedite
the reinforcement learning. Our method uses Large Language Models (LLM) to
obtain high-level domain-specific knowledge using prompt engineering instead of
providing the reinforcement learning algorithm directly with the high-level
knowledge which requires an expert to encode the automaton. We use
chain-of-thought and few-shot methods for prompt engineering and demonstrate
that our method works using these approaches. Additionally, LARL-RM allows for
fully closed-loop reinforcement learning without the need for an expert to
guide and supervise the learning since LARL-RM can use the LLM directly to
generate the required high-level knowledge for the task at hand. We also show
the theoretical guarantee of our algorithm to converge to an optimal policy. We
demonstrate that LARL-RM speeds up the convergence by 30% by implementing our
method in two case studies. | cs.LG | Using Large Language Models to Automate and Expedite Reinforcement Learning with Reward Machine |
2024-02-10T21:46:34Z | http://arxiv.org/pdf/2402.07051v1 | http://arxiv.org/abs/2402.07051v1 | Marcell Vazquez-Chanlatte, Karim Elmaaroufi, Stefan J. Witwicki, Sanjit A. Seshia | 2024-02-10T21:46:34Z | cs.LG, cs.AI, cs.CL, cs.FL | Expert demonstrations have proven an easy way to indirectly specify complex
tasks. Recent algorithms even support extracting unambiguous formal
specifications, e.g. deterministic finite automata (DFA), from demonstrations.
Unfortunately, these techniques are generally not sample efficient. In this
work, we introduce $L^*LM$, an algorithm for learning DFAs from both
demonstrations and natural language. Due to the expressivity of natural
language, we observe a significant improvement in the data efficiency of
learning DFAs from expert demonstrations. Technically, $L^*LM$ leverages large
language models to answer membership queries about the underlying task. This is
then combined with recent techniques for transforming learning from
demonstrations into a sequence of labeled example learning problems. In our
experiments, we observe the two modalities complement each other, yielding a
powerful few-shot learner. | cs.LG | $L^*LM$: Learning Automata from Examples using Natural Language Oracles |
2024-02-10T21:06:34Z | http://arxiv.org/pdf/2402.07043v2 | http://arxiv.org/abs/2402.07043v2 | Elvis Dohmatob, Yunzhen Feng, Pu Yang, Francois Charton, Julia Kempe | 2024-05-31T12:27:52Z | cs.LG, cs.AI, cs.CL | As AI model size grows, neural scaling laws have become a crucial tool to
predict the improvements of large models when increasing capacity and the size
of original (human or natural) training data. Yet, the widespread use of
popular models means that the ecosystem of online data and text will co-evolve
to progressively contain increased amounts of synthesized data. In this paper
we ask: How will the scaling laws change in the inevitable regime where
synthetic data makes its way into the training corpus? Will future models,
still improve, or be doomed to degenerate up to total (model) collapse? We
develop a theoretical framework of model collapse through the lens of scaling
laws. We discover a wide range of decay phenomena, analyzing loss of scaling,
shifted scaling with number of generations, the ''un-learning" of skills, and
grokking when mixing human and synthesized data. Our theory is validated by
large-scale experiments with a transformer on an arithmetic task and text
generation using the large language model Llama2. | cs.LG | A Tale of Tails: Model Collapse as a Change of Scaling Laws |
2024-02-10T19:27:22Z | http://arxiv.org/pdf/2402.07028v1 | http://arxiv.org/abs/2402.07028v1 | Paul Garnier, Gauthier Guinet | 2024-02-10T19:27:22Z | cs.CL, cs.LG | We consider the problem of aligning two sets of continuous word
representations, corresponding to languages, to a common space in order to
infer a bilingual lexicon. It was recently shown that it is possible to infer
such lexicon, without using any parallel data, by aligning word embeddings
trained on monolingual data. Such line of work is called unsupervised bilingual
induction. By wondering whether it was possible to gain experience in the
progressive learning of several languages, we asked ourselves to what extent we
could integrate the knowledge of a given set of languages when learning a new
one, without having parallel data for the latter. In other words, while keeping
the core problem of unsupervised learning in the latest step, we allowed the
access to other corpora of idioms, hence the name semi-supervised. This led us
to propose a novel formulation, considering the lexicon induction as a ranking
problem for which we used recent tools of this machine learning field. Our
experiments on standard benchmarks, inferring dictionary from English to more
than 20 languages, show that our approach consistently outperforms existing
state of the art benchmark. In addition, we deduce from this new scenario
several relevant conclusions allowing a better understanding of the alignment
phenomenon. | cs.CL | Semi-Supervised Learning for Bilingual Lexicon Induction |
2024-02-10T19:08:28Z | http://arxiv.org/pdf/2402.07023v1 | http://arxiv.org/abs/2402.07023v1 | Ankit Pal, Malaikannan Sankarasubbu | 2024-02-10T19:08:28Z | cs.CL, cs.AI, cs.CV, cs.HC, cs.LG | Large language models have the potential to be valuable in the healthcare
industry, but it's crucial to verify their safety and effectiveness through
rigorous evaluation. For this purpose, we comprehensively evaluated both
open-source LLMs and Google's new multimodal LLM called Gemini across Medical
reasoning, hallucination detection, and Medical Visual Question Answering
tasks. While Gemini showed competence, it lagged behind state-of-the-art models
like MedPaLM 2 and GPT-4 in diagnostic accuracy. Additionally, Gemini achieved
an accuracy of 61.45\% on the medical VQA dataset, significantly lower than
GPT-4V's score of 88\%. Our analysis revealed that Gemini is highly susceptible
to hallucinations, overconfidence, and knowledge gaps, which indicate risks if
deployed uncritically. We also performed a detailed analysis by medical subject
and test type, providing actionable feedback for developers and clinicians. To
mitigate risks, we applied prompting strategies that improved performance.
Additionally, we facilitated future research and development by releasing a
Python module for medical LLM evaluation and establishing a dedicated
leaderboard on Hugging Face for medical domain LLMs. Python module can be found
at https://github.com/promptslab/RosettaEval | cs.CL | Gemini Goes to Med School: Exploring the Capabilities of Multimodal Large Language Models on Medical Challenge Problems & Hallucinations |
2024-02-10T16:54:28Z | http://arxiv.org/pdf/2402.06992v1 | http://arxiv.org/abs/2402.06992v1 | Raja Marjieh, Pol van Rijn, Ilia Sucholutsky, Harin Lee, Thomas L. Griffiths, Nori Jacoby | 2024-02-10T16:54:28Z | q-bio.NC, cs.AI, cs.CL, stat.AP | The speech-to-song illusion is a robust psychological phenomenon whereby a
spoken sentence sounds increasingly more musical as it is repeated. Despite
decades of research, a complete formal account of this transformation is still
lacking, and some of its nuanced characteristics, namely, that certain phrases
appear to transform while others do not, is not well understood. Here we
provide a formal account of this phenomenon, by recasting it as a statistical
inference whereby a rational agent attempts to decide whether a sequence of
utterances is more likely to have been produced in a song or speech. Using this
approach and analyzing song and speech corpora, we further introduce a novel
prose-to-lyrics illusion that is purely text-based. In this illusion, simply
duplicating written sentences makes them appear more like song lyrics. We
provide robust evidence for this new illusion in both human participants and
large language models. | q-bio.NC | A Rational Analysis of the Speech-to-Song Illusion |
2024-02-10T16:48:45Z | http://arxiv.org/pdf/2402.10951v1 | http://arxiv.org/abs/2402.10951v1 | Chris von Csefalvay | 2024-02-10T16:48:45Z | cs.CL, cs.LG, 68T50, I.2.7; J.3 | Over the recent years, the emergence of large language models (LLMs) has
given rise to a proliferation of domain-specific models that are intended to
reflect the particularities of linguistic context and content as a correlate of
the originating domain. This paper details the conception, design, training and
evaluation of DAEDRA, a LLM designed to detect regulatory-relevant outcomes
(mortality, ER attendance and hospitalisation) in adverse event reports
elicited through passive reporting (PR). While PR is a highly cost-efficient
way of eliciting information from a wide and diverse audience -- typically
including not only physicians and healthcare providers but also patients,
family members and other lay stakeholders --, this diversity makes PR corpora
difficult to analyse. Generic language models may not capture the complex
clinical dimensions while specific clinical or biomedical models may not
perform well on lay reports. To evaluate the utility of a subdomain-specific
language model, an adaptive training approach was adapted, wherein base
language model candidates were evaluated on a subset of the corpus, and the
best performer was trained on the entire corpus. This yielded a small but
significant improvement in $F_1$ (+1%), precision (+2.5%) and recall (+3.8%),
at a relatively low training cost and a single-day training time.
Subdomain-specific LLMs continue to be viable options for better results when
analysing highly specialised corpora. | cs.CL | DAEDRA: A language model for predicting outcomes in passive pharmacovigilance reporting |
2024-02-10T15:32:53Z | http://arxiv.org/pdf/2402.06973v1 | http://arxiv.org/abs/2402.06973v1 | William Gantt, Alexander Martin, Pavlo Kuchmiichuk, Aaron Steven White | 2024-02-10T15:32:53Z | cs.CL, cs.AI, cs.LG | We introduce event-keyed summarization (EKS), a novel task that marries
traditional summarization and document-level event extraction, with the goal of
generating a contextualized summary for a specific event, given a document and
an extracted event structure. We introduce a dataset for this task, MUCSUM,
consisting of summaries of all events in the classic MUC-4 dataset, along with
a set of baselines that comprises both pretrained LM standards in the
summarization literature, as well as larger frontier models. We show that
ablations that reduce EKS to traditional summarization or structure-to-text
yield inferior summaries of target events and that MUCSUM is a robust benchmark
for this task. Lastly, we conduct a human evaluation of both reference and
model summaries, and provide some detailed analysis of the results. | cs.CL | Event-Keyed Summarization |
2024-02-10T14:52:52Z | http://arxiv.org/pdf/2402.06967v2 | http://arxiv.org/abs/2402.06967v2 | Jian Wang, Chak Tou Leong, Jiashuo Wang, Dongding Lin, Wenjie Li, Xiao-Yong Wei | 2024-05-30T04:57:36Z | cs.CL, cs.AI | Tuning language models for dialogue generation has been a prevalent paradigm
for building capable dialogue agents. Yet, traditional tuning narrowly views
dialogue generation as resembling other language generation tasks, ignoring the
role disparities between two speakers and the multi-round interactive process
that dialogues ought to be. Such a manner often leads to unsatisfactory chat
consistency for the built agent. In this work, we emphasize the interactive,
communicative nature of dialogue and argue that it is more feasible to model
the speaker roles of agent and user separately, enabling the agent to adhere to
its role consistently. With this in mind, we propose an efficient Multi-round
Interactive Dialogue Tuning (Midi-Tuning) framework. It models the agent and
user individually with two adapters built upon large language models. The
adapters make use of respective utterances round by round in alternating order
and they are tuned via a round-level memory caching mechanism. Extensive
experiments demonstrate that, our framework performs superior to traditional
fine-tuning and harbors the tremendous potential for improving dialogue
consistency. | cs.CL | Instruct Once, Chat Consistently in Multiple Rounds: An Efficient Tuning Framework for Dialogue |
2024-02-10T14:43:08Z | http://arxiv.org/pdf/2402.06964v1 | http://arxiv.org/abs/2402.06964v1 | Francis G. VanGessel, Efrem Perry, Salil Mohan, Oliver M. Barham, Mark Cavolowsky | 2024-02-10T14:43:08Z | cs.CL, cond-mat.mtrl-sci | We present a demonstration of the utility of NLP for aiding research into
energetic materials and associated systems. The NLP method enables machine
understanding of textual data, offering an automated route to knowledge
discovery and information extraction from energetics text. We apply three
established unsupervised NLP models: Latent Dirichlet Allocation, Word2Vec, and
the Transformer to a large curated dataset of energetics-related scientific
articles. We demonstrate that each NLP algorithm is capable of identifying
energetic topics and concepts, generating a language model which aligns with
Subject Matter Expert knowledge. Furthermore, we present a document
classification pipeline for energetics text. Our classification pipeline
achieves 59-76\% accuracy depending on the NLP model used, with the highest
performing Transformer model rivaling inter-annotator agreement metrics. The
NLP approaches studied in this work can identify concepts germane to energetics
and therefore hold promise as a tool for accelerating energetics research
efforts and energetics material development. | cs.CL | NLP for Knowledge Discovery and Information Extraction from Energetics Corpora |
2024-02-10T14:26:42Z | http://arxiv.org/pdf/2402.06959v1 | http://arxiv.org/abs/2402.06959v1 | Hsuan-Fu Wang, Yi-Jen Shih, Heng-Jui Chang, Layne Berry, Puyuan Peng, Hung-yi Lee, Hsin-Min Wang, David Harwath | 2024-02-10T14:26:42Z | cs.CL, cs.SD, eess.AS | The recently proposed visually grounded speech model SpeechCLIP is an
innovative framework that bridges speech and text through images via CLIP
without relying on text transcription. On this basis, this paper introduces two
extensions to SpeechCLIP. First, we apply the Continuous Integrate-and-Fire
(CIF) module to replace a fixed number of CLS tokens in the cascaded
architecture. Second, we propose a new hybrid architecture that merges the
cascaded and parallel architectures of SpeechCLIP into a multi-task learning
framework. Our experimental evaluation is performed on the Flickr8k and
SpokenCOCO datasets. The results show that in the speech keyword extraction
task, the CIF-based cascaded SpeechCLIP model outperforms the previous cascaded
SpeechCLIP model using a fixed number of CLS tokens. Furthermore, through our
hybrid architecture, cascaded task learning boosts the performance of the
parallel branch in image-speech retrieval tasks. | cs.CL | SpeechCLIP+: Self-supervised multi-task representation learning for speech via CLIP and speech-image data |
2024-02-10T13:50:11Z | http://arxiv.org/pdf/2402.06954v1 | http://arxiv.org/abs/2402.06954v1 | Rui Ye, Wenhao Wang, Jingyi Chai, Dihan Li, Zexi Li, Yinda Xu, Yaxin Du, Yanfeng Wang, Siheng Chen | 2024-02-10T13:50:11Z | cs.LG, cs.CL, cs.DC, cs.MA | Trained on massive publicly available data, large language models (LLMs) have
demonstrated tremendous success across various fields. While more data
contributes to better performance, a disconcerting reality is that high-quality
public data will be exhausted in a few years. In this paper, we offer a
potential next step for contemporary LLMs: collaborative and privacy-preserving
LLM training on the underutilized distributed private data via federated
learning (FL), where multiple data owners collaboratively train a shared model
without transmitting raw data. To achieve this, we build a concise, integrated,
and research-friendly framework/codebase, named OpenFedLLM. It covers federated
instruction tuning for enhancing instruction-following capability, federated
value alignment for aligning with human values, and 7 representative FL
algorithms. Besides, OpenFedLLM supports training on diverse domains, where we
cover 8 training datasets; and provides comprehensive evaluations, where we
cover 30+ evaluation metrics. Through extensive experiments, we observe that
all FL algorithms outperform local training on training LLMs, demonstrating a
clear performance improvement across a variety of settings. Notably, in a
financial benchmark, Llama2-7B fine-tuned by applying any FL algorithm can
outperform GPT-4 by a significant margin while the model obtained through
individual training cannot, demonstrating strong motivation for clients to
participate in FL. The code is available at
https://github.com/rui-ye/OpenFedLLM. | cs.LG | OpenFedLLM: Training Large Language Models on Decentralized Private Data via Federated Learning |
2024-02-10T13:26:14Z | http://arxiv.org/pdf/2402.06948v1 | http://arxiv.org/abs/2402.06948v1 | Nefeli Gkouti, Prodromos Malakasiotis, Stavros Toumpis, Ion Androutsopoulos | 2024-02-10T13:26:14Z | cs.CL | NLP research has explored different neural model architectures and sizes,
datasets, training objectives, and transfer learning techniques. However, the
choice of optimizer during training has not been explored as extensively.
Typically, some variant of Stochastic Gradient Descent (SGD) is employed,
selected among numerous variants, using unclear criteria, often with minimal or
no tuning of the optimizer's hyperparameters. Experimenting with five GLUE
datasets, two models (DistilBERT and DistilRoBERTa), and seven popular
optimizers (SGD, SGD with Momentum, Adam, AdaMax, Nadam, AdamW, and AdaBound),
we find that when the hyperparameters of the optimizers are tuned, there is no
substantial difference in test performance across the five more elaborate
(adaptive) optimizers, despite differences in training loss. Furthermore,
tuning just the learning rate is in most cases as good as tuning all the
hyperparameters. Hence, we recommend picking any of the best-behaved adaptive
optimizers (e.g., Adam) and tuning only its learning rate. When no
hyperparameter can be tuned, SGD with Momentum is the best choice. | cs.CL | Should I try multiple optimizers when fine-tuning pre-trained Transformers for NLP tasks? Should I tune their hyperparameters? |
2024-02-10T11:53:48Z | http://arxiv.org/pdf/2402.06930v1 | http://arxiv.org/abs/2402.06930v1 | Chufan Shi, Deng Cai, Yujiu Yang | 2024-02-10T11:53:48Z | cs.CL | In the rapidly evolving field of text generation, the demand for more precise
control mechanisms has become increasingly apparent. To address this need, we
present a novel methodology, LIFI, which offers a lightweight approach with
fine-grained control for controlled text generation. Unlike previous studies
that train pre-trained language models to follow discrete, categorical, and
exclusive control codes, LIFI learns controlled text generation under the
guidance of continuous, relative, and nonexclusive control codes. These
fine-grained codes are automatically derived from an attribute classifier,
initially trained with a small amount of labeled data and subsequently employed
to label abundant unlabeled data, thus garnering more extensive supervision
signals. Moreover, to achieve efficient control, we incorporate the
fine-grained control codes with adapters, a parameter- and compute-efficient
way to steer a pre-trained language model. We evaluate LIFI on two conventional
tasks -- sentiment control and topic control -- and one newly proposed task --
stylistic novel writing. Comprehensive experimental results validate the
effectiveness of our proposed methods, demonstrating substantial performance
improvements over existing baselines. | cs.CL | LiFi: Lightweight Controlled Text Generation with Fine-Grained Control Codes |
2024-02-10T11:14:53Z | http://arxiv.org/pdf/2402.06925v2 | http://arxiv.org/abs/2402.06925v2 | Chufan Shi, Haoran Yang, Deng Cai, Zhisong Zhang, Yifan Wang, Yujiu Yang, Wai Lam | 2024-06-17T14:34:05Z | cs.CL | Decoding methods play an indispensable role in converting language models
from next-token predictors into practical task solvers. Prior research on
decoding methods, primarily focusing on task-specific models, may not extend to
the current era of general-purpose large language models (LLMs). Moreover, the
recent influx of decoding strategies has further complicated this landscape.
This paper provides a comprehensive and multifaceted analysis of various
decoding methods within the context of LLMs, evaluating their performance,
robustness to hyperparameter changes, and decoding speeds across a wide range
of tasks, models, and deployment environments. Our findings reveal that
decoding method performance is notably task-dependent and influenced by factors
such as alignment, model size, and quantization. Intriguingly, sensitivity
analysis exposes that certain methods achieve superior performance at the cost
of extensive hyperparameter tuning, highlighting the trade-off between
attaining optimal results and the practicality of implementation in varying
contexts. | cs.CL | A Thorough Examination of Decoding Methods in the Era of LLMs |
2024-02-10T09:51:03Z | http://arxiv.org/pdf/2402.06918v2 | http://arxiv.org/abs/2402.06918v2 | Zhen-Yu Zhang, Siwei Han, Huaxiu Yao, Gang Niu, Masashi Sugiyama | 2024-06-26T05:47:52Z | cs.LG, cs.AI, cs.CL | To improve the ability of the large language model (LLMs) to tackle complex
reasoning problems, chain-of-thoughts (CoT) methods were proposed to guide LLMs
to reason step-by-step, enabling problem solving from simple to complex.
State-of-the-art methods for generating such a chain involve interactive
collaboration, where the learner generates candidate intermediate thoughts,
evaluated by the LLM, guiding the generation of subsequent thoughts. However, a
widespread yet understudied problem is that the evaluation from the LLM is
typically noisy and unreliable, potentially misleading the generation process
in selecting promising intermediate thoughts. In this paper, motivated by
Vapnik's principle, we use pairwise-comparison evaluation instead of point-wise
scoring to search for promising intermediate thoughts with the noisy feedback
from the LLM. In each round, we randomly pair intermediate thoughts and
directly prompt the LLM to select the more promising one from each pair,
allowing us to identify the most promising thoughts through an iterative
process. To further alleviate the noise in the comparison, we incorporate
techniques from ensemble learning and dueling bandits, proposing two variants
of the algorithm. Experiments on three real-world tasks demonstrate the
effectiveness of our proposed algorithm and verify the rationale of the
pairwise comparison mechanism. | cs.LG | Generating Chain-of-Thoughts with a Pairwise-Comparison Approach to Searching for the Most Promising Intermediate Thought |
2024-02-10T09:16:56Z | http://arxiv.org/pdf/2402.06913v1 | http://arxiv.org/abs/2402.06913v1 | Shahbaz Syed, Khalid Al-Khatib, Martin Potthast | 2024-02-10T09:16:56Z | cs.CL | This paper presents TL;DR Progress, a new tool for exploring the literature
on neural text summarization. It organizes 514~papers based on a comprehensive
annotation scheme for text summarization approaches and enables fine-grained,
faceted search. Each paper was manually annotated to capture aspects such as
evaluation metrics, quality dimensions, learning paradigms, challenges
addressed, datasets, and document domains. In addition, a succinct indicative
summary is provided for each paper, consisting of automatically extracted
contextual factors, issues, and proposed solutions. The tool is available
online at https://www.tldr-progress.de, a demo video at
https://youtu.be/uCVRGFvXUj8 | cs.CL | TL;DR Progress: Multi-faceted Literature Exploration in Text Summarization |
2024-02-10T08:25:30Z | http://arxiv.org/pdf/2402.06907v1 | http://arxiv.org/abs/2402.06907v1 | Chen Jia-Chen, Guillem Senabre, Allane Caron | 2024-02-10T08:25:30Z | cs.CL | With more and more advanced data analysis techniques emerging, people will
expect these techniques to be applied in more complex tasks and solve problems
in our daily lives. Text Summarization is one of famous applications in Natural
Language Processing (NLP) field. It aims to automatically generate summary with
important information based on a given context, which is important when you
have to deal with piles of documents. Summarization techniques can help capture
key points in a short time and bring convenience in works. One of applicable
situation is meeting summarization, especially for important meeting that tend
to be long, complicated, multi-topic and multi-person. Therefore, when people
want to review specific content from a meeting, it will be hard and
time-consuming to find the related spans in the meeting transcript. However,
most of previous works focus on doing summarization for newsletters, scientific
articles...etc, which have a clear document structure and an official format.
For the documents with complex structure like transcripts, we think those works
are not quite suitable for meeting summarization. Besides, the consistency of
summary is another issue common to be discussed in NLP field. To conquer
challenges of meeting summarization, we are inspired by "QMSum: A New Benchmark
for Query-based Multi-domain Meeting Summarization" proposed by Microsoft and
we also propose our Locater model designed to extract relevant spans based on
given transcript and query, which are then summarized by Summarizer model.
Furthermore, we perform a comparative study by applying different word
embedding techniques to improve summary consistency. | cs.CL | Investigating Consistency in Query-Based Meeting Summarization: A Comparative Study of Different Embedding Methods |
2024-02-10T08:24:29Z | http://arxiv.org/pdf/2403.05553v1 | http://arxiv.org/abs/2403.05553v1 | Tamador Alkhidir, Edmond Awad, Aamena Alshamsi | 2024-02-10T08:24:29Z | cs.CY, cs.CL, cs.LG | Education systems are dynamically changing to accommodate technological
advances, industrial and societal needs, and to enhance students' learning
journeys. Curriculum specialists and educators constantly revise taught
subjects across educational grades to identify gaps, introduce new learning
topics, and enhance the learning outcomes. This process is usually done within
the same subjects (e.g. math) or across related subjects (e.g. math and
physics) considering the same and different educational levels, leading to
massive multi-layer comparisons. Having nuanced data about subjects, topics,
and learning outcomes structured within a dataset, empowers us to leverage data
science to better understand the progression of various learning topics. In
this paper, Bidirectional Encoder Representations from Transformers (BERT)
topic modeling was used to extract topics from the curriculum, which were then
used to identify relationships between subjects, track their progression, and
identify conceptual gaps. We found that grouping learning outcomes by common
topics helped specialists reduce redundancy and introduce new concepts in the
curriculum. We built a dashboard to avail the methodology to curriculum
specials. Finally, we tested the validity of the approach with subject matter
experts. | cs.CY | Understanding the Progression of Educational Topics via Semantic Matching |
2024-02-10T07:55:27Z | http://arxiv.org/pdf/2402.06900v3 | http://arxiv.org/abs/2402.06900v3 | Hyukhun Koh, Dohyung Kim, Minwoo Lee, Kyomin Jung | 2024-06-18T04:35:12Z | cs.CL, cs.AI | In the pursuit of developing Large Language Models (LLMs) that adhere to
societal standards, it is imperative to detect the toxicity in the generated
text. The majority of existing toxicity metrics rely on encoder models trained
on specific toxicity datasets, which are susceptible to out-of-distribution
(OOD) problems and depend on the dataset's definition of toxicity. In this
paper, we introduce a robust metric grounded on LLMs to flexibly measure
toxicity according to the given definition. We first analyze the toxicity
factors, followed by an examination of the intrinsic toxic attributes of LLMs
to ascertain their suitability as evaluators. Finally, we evaluate the
performance of our metric with detailed analysis. Our empirical results
demonstrate outstanding performance in measuring toxicity within verified
factors, improving on conventional metrics by 12 points in the F1 score. Our
findings also indicate that upstream toxicity significantly influences
downstream metrics, suggesting that LLMs are unsuitable for toxicity
evaluations within unverified factors. | cs.CL | Can LLMs Recognize Toxicity? Definition-Based Toxicity Metric |
2024-02-10T07:20:49Z | http://arxiv.org/pdf/2402.06894v2 | http://arxiv.org/abs/2402.06894v2 | Yuchen Hu, Chen Chen, Chao-Han Huck Yang, Ruizhe Li, Dong Zhang, Zhehuai Chen, Eng Siong Chng | 2024-05-16T13:17:05Z | cs.CL, cs.AI, cs.LG, cs.SD, eess.AS | Recent advances in large language models (LLMs) have stepped forward the
development of multilingual speech and machine translation by its reduced
representation errors and incorporated external knowledge. However, both
translation tasks typically utilize beam search decoding and top-1 hypothesis
selection for inference. These techniques struggle to fully exploit the rich
information in the diverse N-best hypotheses, making them less optimal for
translation tasks that require a single, high-quality output sequence. In this
paper, we propose a new generative paradigm for translation tasks, namely
"GenTranslate", which builds upon LLMs to generate better results from the
diverse translation versions in N-best list. Leveraging the rich linguistic
knowledge and strong reasoning abilities of LLMs, our new paradigm can
integrate the rich information in N-best candidates to generate a
higher-quality translation result. Furthermore, to support LLM finetuning, we
build and release a HypoTranslate dataset that contains over 592K
hypotheses-translation pairs in 11 languages. Experiments on various speech and
machine translation benchmarks (e.g., FLEURS, CoVoST-2, WMT) demonstrate that
our GenTranslate significantly outperforms the state-of-the-art model. | cs.CL | GenTranslate: Large Language Models are Generative Multilingual Speech and Machine Translators |
2024-02-10T04:47:58Z | http://arxiv.org/pdf/2402.07950v1 | http://arxiv.org/abs/2402.07950v1 | Shariq Murtuza | 2024-02-10T04:47:58Z | cs.CR, cs.CL | With the release of OpenAI's ChatGPT, the field of large language models
(LLM) saw an increase of academic interest in GPT based chat assistants. In the
next few months multiple accesible large language models were released that
included Meta's LLama models and Mistral AI's Mistral and Mixtral MoE models.
These models are available openly for a wide array of purposes with a wide
spectrum of licenses. These LLMs have found their use in a different number of
fields like code development, SQL generation etc. In this work we propose our
plan to explore the applicability of large language model in the domain of
network security. We plan to create Sentinel, a LLM, to analyse network packet
contents and pass a judgment on it's threat level. This work is a preliminary
report that will lay our plan for our future endeavors. | cs.CR | Sentinels of the Stream: Unleashing Large Language Models for Dynamic Packet Classification in Software Defined Networks -- Position Paper |
2024-02-10T01:18:15Z | http://arxiv.org/pdf/2402.06853v2 | http://arxiv.org/abs/2402.06853v2 | Zhibo Chu, Shiwen Ni, Zichong Wang, Xi Feng, Min Yang, Wenbin Zhang | 2024-06-12T02:28:12Z | cs.CL | Language models serve as a cornerstone in natural language processing (NLP),
utilizing mathematical methods to generalize language laws and knowledge for
prediction and generation. Over extensive research spanning decades, language
modeling has progressed from initial statistical language models (SLMs) to the
contemporary landscape of large language models (LLMs). Notably, the swift
evolution of LLMs has reached the ability to process, understand, and generate
human-level text. Nevertheless, despite the significant advantages that LLMs
offer in improving both work and personal lives, the limited understanding
among general practitioners about the background and principles of these models
hampers their full potential. Notably, most LLMs reviews focus on specific
aspects and utilize specialized language, posing a challenge for practitioners
lacking relevant background knowledge. In light of this, this survey aims to
present a comprehensible overview of LLMs to assist a broader audience. It
strives to facilitate a comprehensive understanding by exploring the historical
background of language models and tracing their evolution over time. The survey
further investigates the factors influencing the development of LLMs,
emphasizing key contributions. Additionally, it concentrates on elucidating the
underlying principles of LLMs, equipping audiences with essential theoretical
knowledge. The survey also highlights the limitations of existing work and
points out promising future directions. | cs.CL | History, Development, and Principles of Large Language Models-An Introductory Survey |
2024-02-10T01:11:59Z | http://arxiv.org/pdf/2402.06852v2 | http://arxiv.org/abs/2402.06852v2 | Di Zhang, Wei Liu, Qian Tan, Jingdan Chen, Hang Yan, Yuliang Yan, Jiatong Li, Weiran Huang, Xiangyu Yue, Wanli Ouyang, Dongzhan Zhou, Shufei Zhang, Mao Su, Han-Sen Zhong, Yuqiang Li | 2024-04-25T14:34:28Z | cs.AI, cs.CL | Large language models (LLMs) have made impressive progress in chemistry
applications. However, the community lacks an LLM specifically designed for
chemistry. The main challenges are two-fold: firstly, most chemical data and
scientific knowledge are stored in structured databases, which limits the
model's ability to sustain coherent dialogue when used directly. Secondly,
there is an absence of objective and fair benchmark that encompass most
chemistry tasks. Here, we introduce ChemLLM, a comprehensive framework that
features the first LLM dedicated to chemistry. It also includes ChemData, a
dataset specifically designed for instruction tuning, and ChemBench, a robust
benchmark covering nine essential chemistry tasks. ChemLLM is adept at
performing various tasks across chemical disciplines with fluid dialogue
interaction. Notably, ChemLLM achieves results comparable to GPT-4 on the core
chemical tasks and demonstrates competitive performance with LLMs of similar
size in general scenarios. ChemLLM paves a new path for exploration in chemical
studies, and our method of incorporating structured chemical knowledge into
dialogue systems sets a new standard for developing LLMs in various scientific
fields. Codes, Datasets, and Model weights are publicly accessible at
https://hf.co/AI4Chem | cs.AI | ChemLLM: A Chemical Large Language Model |
2024-02-09T23:02:57Z | http://arxiv.org/pdf/2402.06820v2 | http://arxiv.org/abs/2402.06820v2 | Tiago Mendes-Neves, Luís Meireles, João Mendes-Moreira | 2024-04-26T11:45:02Z | cs.LG, cs.CL | This paper introduces an approach to predicting the next event in a soccer
match, a challenge bearing remarkable similarities to the problem faced by
Large Language Models (LLMs). Unlike other methods that severely limit event
dynamics in soccer, often abstracting from many variables or relying on a mix
of sequential models, our research proposes a novel technique inspired by the
methodologies used in LLMs. These models predict a complete chain of variables
that compose an event, significantly simplifying the construction of Large
Event Models (LEMs) for soccer. Utilizing deep learning on the publicly
available WyScout dataset, the proposed approach notably surpasses the
performance of previous LEM proposals in critical areas, such as the prediction
accuracy of the next event type. This paper highlights the utility of LEMs in
various applications, including match prediction and analytics. Moreover, we
show that LEMs provide a simulation backbone for users to build many analytics
pipelines, an approach opposite to the current specialized single-purpose
models. LEMs represent a pivotal advancement in soccer analytics, establishing
a foundational framework for multifaceted analytics pipelines through a
singular machine-learning model. | cs.LG | Forecasting Events in Soccer Matches Through Language |
2024-02-09T22:48:45Z | http://arxiv.org/pdf/2402.10949v2 | http://arxiv.org/abs/2402.10949v2 | Rick Battle, Teja Gollapudi | 2024-02-20T15:03:00Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) have demonstrated remarkable problem-solving and
basic mathematics abilities. However, their efficacy is highly contingent on
the formulation of the prompt. This study endeavors to quantify the influence
of incorporating "positive thinking" into the system message of the prompt,
then compare that to systematic prompt optimization. We assess the performance
of 60 combinations of system message snippets, tested with and without Chain of
Thought prompting, across three models with parameters ranging from 7 to 70
billion on the GSM8K dataset. Our findings reveal that results do not
universally generalize across models. In most instances, the inclusion of
"positive thinking" prompts positively affected model performance. Notably,
however, Llama2-70B exhibited an exception when not utilizing Chain of Thought,
as the optimal system message was found to be none at all. Given the
combinatorial complexity, and thus computation time, of experimenting with
hand-tuning prompts for large black-box models, we then compared the
performance of the best "positive thinking" prompt against the output of
systematic prompt optimization. We show that employing an automated prompt
optimizer emerges as the most effective method for enhancing performance, even
when working with smaller open-source models. Additionally, our findings reveal
that the highest-scoring, automatically-optimized prompt exhibits a degree of
peculiarity far beyond expectations. | cs.CL | The Unreasonable Effectiveness of Eccentric Automatic Prompts |
2024-02-09T22:12:20Z | http://arxiv.org/pdf/2403.15393v1 | http://arxiv.org/abs/2403.15393v1 | Yuchen Wang, Zhengyu Fang, Wei Du, Shuai Xu, Rong Xu, Jing Li | 2024-02-09T22:12:20Z | cs.CL, cs.LG, cs.SI | The opioid epidemic, referring to the growing hospitalizations and deaths
because of overdose of opioid usage and addiction, has become a severe health
problem in the United States. Many strategies have been developed by the
federal and local governments and health communities to combat this crisis.
Among them, improving our understanding of the epidemic through better health
surveillance is one of the top priorities. In addition to direct testing,
machine learning approaches may also allow us to detect opioid users by
analyzing data from social media because many opioid users may choose not to do
the tests but may share their experiences on social media anonymously. In this
paper, we take advantage of recent advances in machine learning, collect and
analyze user posts from a popular social network Reddit with the goal to
identify opioid users. Posts from more than 1,000 users who have posted on
three sub-reddits over a period of one month have been collected. In addition
to the ones that contain keywords such as opioid, opiate, or heroin, we have
also collected posts that contain slang words of opioid such as black or
chocolate. We apply an attention-based bidirectional long short memory model to
identify opioid users. Experimental results show that the approaches
significantly outperform competitive algorithms in terms of F1-score.
Furthermore, the model allows us to extract most informative words, such as
opiate, opioid, and black, from posts via the attention layer, which provides
more insights on how the machine learning algorithm works in distinguishing
drug users from non-drug users. | cs.CL | Detection of Opioid Users from Reddit Posts via an Attention-based Bidirectional Recurrent Neural Network |
2024-02-09T21:05:01Z | http://arxiv.org/pdf/2402.06782v3 | http://arxiv.org/abs/2402.06782v3 | Akbir Khan, John Hughes, Dan Valentine, Laura Ruis, Kshitij Sachan, Ansh Radhakrishnan, Edward Grefenstette, Samuel R. Bowman, Tim Rocktäschel, Ethan Perez | 2024-05-30T13:59:34Z | cs.AI, cs.CL | Common methods for aligning large language models (LLMs) with desired
behaviour heavily rely on human-labelled data. However, as models grow
increasingly sophisticated, they will surpass human expertise, and the role of
human evaluation will evolve into non-experts overseeing experts. In
anticipation of this, we ask: can weaker models assess the correctness of
stronger models? We investigate this question in an analogous setting, where
stronger models (experts) possess the necessary information to answer questions
and weaker models (non-experts) lack this information. The method we evaluate
is debate, where two LLM experts each argue for a different answer, and a
non-expert selects the answer. We find that debate consistently helps both
non-expert models and humans answer questions, achieving 76% and 88% accuracy
respectively (naive baselines obtain 48% and 60%). Furthermore, optimising
expert debaters for persuasiveness in an unsupervised manner improves
non-expert ability to identify the truth in debates. Our results provide
encouraging empirical evidence for the viability of aligning models with debate
in the absence of ground truth. | cs.AI | Debating with More Persuasive LLMs Leads to More Truthful Answers |
2024-02-09T19:59:34Z | http://arxiv.org/pdf/2402.06766v1 | http://arxiv.org/abs/2402.06766v1 | Marcellus Amadeus, William Alberto Cruz Castañeda | 2024-02-09T19:59:34Z | cs.CL, cs.AI | Recent surveys on data augmentation for natural language processing have
reported different techniques and advancements in the field. Several
frameworks, tools, and repositories promote the implementation of text data
augmentation pipelines. However, a lack of evaluation criteria and standards
for method comparison due to different tasks, metrics, datasets, architectures,
and experimental settings makes comparisons meaningless. Also, a lack of
methods unification exists and text data augmentation research would benefit
from unified metrics to compare different augmentation methods. Thus, academics
and the industry endeavor relevant evaluation metrics for text data
augmentation techniques. The contribution of this work is to provide a taxonomy
of evaluation metrics for text augmentation methods and serve as a direction
for a unified benchmark. The proposed taxonomy organizes categories that
include tools for implementation and metrics calculation. Finally, with this
study, we intend to present opportunities to explore the unification and
standardization of text data augmentation metrics. | cs.CL | Evaluation Metrics for Text Data Augmentation in NLP |
2024-02-09T19:16:27Z | http://arxiv.org/pdf/2402.06738v1 | http://arxiv.org/abs/2402.06738v1 | Yifan Ding, Amrit Poudel, Qingkai Zeng, Tim Weninger, Balaji Veeramani, Sanmitra Bhattacharya | 2024-02-09T19:16:27Z | cs.CL | The ability of Large Language Models (LLMs) to generate factually correct
output remains relatively unexplored due to the lack of fact-checking and
knowledge grounding during training and inference. In this work, we aim to
address this challenge through the Entity Disambiguation (ED) task. We first
consider prompt engineering, and design a three-step hard-prompting method to
probe LLMs' ED performance without supervised fine-tuning (SFT). Overall, the
prompting method improves the micro-F_1 score of the original vanilla models by
a large margin, on some cases up to 36% and higher, and obtains comparable
performance across 10 datasets when compared to existing methods with SFT. We
further improve the knowledge grounding ability through instruction tuning (IT)
with similar prompts and responses. The instruction-tuned model not only
achieves higher micro-F1 score performance as compared to several baseline
methods on supervised entity disambiguation tasks with an average micro-F_1
improvement of 2.1% over the existing baseline models, but also obtains higher
accuracy on six Question Answering (QA) tasks in the zero-shot setting. Our
methodologies apply to both open- and closed-source LLMs. | cs.CL | EntGPT: Linking Generative Large Language Models with Knowledge Bases |
2024-02-09T19:09:19Z | http://arxiv.org/pdf/2402.06733v3 | http://arxiv.org/abs/2402.06733v3 | Pragya Srivastava, Satvik Golechha, Amit Deshpande, Amit Sharma | 2024-06-06T12:16:55Z | cs.CL, cs.AI, cs.LG | Recent work shows that in-context learning and optimization of in-context
examples (ICE) can significantly improve the accuracy of large language models
(LLMs) on a wide range of tasks, leading to an apparent consensus that ICE
optimization is crucial for better performance. However, most of these studies
assume a fixed or no instruction provided in the prompt. We challenge this
consensus by investigating the necessity of optimizing ICE when task-specific
instructions are provided and find that there are many tasks for which it
yields diminishing returns. In particular, using a diverse set of tasks and a
systematically created instruction set with gradually added details, we find
that as the prompt instruction becomes more detailed, the returns on ICE
optimization diminish. To characterize this behavior, we introduce a
task-specific metric called Normalized Invariability to Choice of Examples
(NICE) that quantifies the learnability of tasks from a given instruction, and
provides a heuristic to help decide whether to optimize instructions or ICE for
a new task. Given a task, the proposed metric can reliably predict the utility
of optimizing ICE compared to using random ICE. Our code is available at
https://github.com/microsoft/nice-icl. | cs.CL | NICE: To Optimize In-Context Examples or Not? |
2024-02-09T18:59:29Z | http://arxiv.org/pdf/2402.06627v3 | http://arxiv.org/abs/2402.06627v3 | Alexander Pan, Erik Jones, Meena Jagadeesan, Jacob Steinhardt | 2024-06-06T21:39:09Z | cs.LG, cs.AI, cs.CL | Language models influence the external world: they query APIs that read and
write to web pages, generate content that shapes human behavior, and run system
commands as autonomous agents. These interactions form feedback loops: LLM
outputs affect the world, which in turn affect subsequent LLM outputs. In this
work, we show that feedback loops can cause in-context reward hacking (ICRH),
where the LLM at test-time optimizes a (potentially implicit) objective but
creates negative side effects in the process. For example, consider an LLM
agent deployed to increase Twitter engagement; the LLM may retrieve its
previous tweets into the context window and make them more controversial,
increasing engagement but also toxicity. We identify and study two processes
that lead to ICRH: output-refinement and policy-refinement. For these
processes, evaluations on static datasets are insufficient -- they miss the
feedback effects and thus cannot capture the most harmful behavior. In
response, we provide three recommendations for evaluation to capture more
instances of ICRH. As AI development accelerates, the effects of feedback loops
will proliferate, increasing the need to understand their role in shaping LLM
behavior. | cs.LG | Feedback Loops With Language Models Drive In-Context Reward Hacking |
2024-02-09T18:57:08Z | http://arxiv.org/pdf/2402.06625v1 | http://arxiv.org/abs/2402.06625v1 | Satyapriya Krishna, Chirag Agarwal, Himabindu Lakkaraju | 2024-02-09T18:57:08Z | cs.CL | The development of Large Language Models (LLMs) has notably transformed
numerous sectors, offering impressive text generation capabilities. Yet, the
reliability and truthfulness of these models remain pressing concerns. To this
end, we investigate iterative prompting, a strategy hypothesized to refine LLM
responses, assessing its impact on LLM truthfulness, an area which has not been
thoroughly explored. Our extensive experiments delve into the intricacies of
iterative prompting variants, examining their influence on the accuracy and
calibration of model responses. Our findings reveal that naive prompting
methods significantly undermine truthfulness, leading to exacerbated
calibration errors. In response to these challenges, we introduce several
prompting variants designed to address the identified issues. These variants
demonstrate marked improvements over existing baselines, signaling a promising
direction for future research. Our work provides a nuanced understanding of
iterative prompting and introduces novel approaches to enhance the truthfulness
of LLMs, thereby contributing to the development of more accurate and
trustworthy AI systems. | cs.CL | Understanding the Effects of Iterative Prompting on Truthfulness |
2024-02-09T18:51:49Z | http://arxiv.org/pdf/2402.06619v1 | http://arxiv.org/abs/2402.06619v1 | Shivalika Singh, Freddie Vargus, Daniel Dsouza, Börje F. Karlsson, Abinaya Mahendiran, Wei-Yin Ko, Herumb Shandilya, Jay Patel, Deividas Mataciunas, Laura OMahony, Mike Zhang, Ramith Hettiarachchi, Joseph Wilson, Marina Machado, Luisa Souza Moura, Dominik Krzemiński, Hakimeh Fadaei, Irem Ergün, Ifeoma Okoh, Aisha Alaagib, Oshan Mudannayake, Zaid Alyafeai, Vu Minh Chien, Sebastian Ruder, Surya Guthikonda, Emad A. Alghamdi, Sebastian Gehrmann, Niklas Muennighoff, Max Bartolo, Julia Kreutzer, Ahmet Üstün, Marzieh Fadaee, Sara Hooker | 2024-02-09T18:51:49Z | cs.CL, cs.AI | Datasets are foundational to many breakthroughs in modern artificial
intelligence. Many recent achievements in the space of natural language
processing (NLP) can be attributed to the finetuning of pre-trained models on a
diverse set of tasks that enables a large language model (LLM) to respond to
instructions. Instruction fine-tuning (IFT) requires specifically constructed
and annotated datasets. However, existing datasets are almost all in the
English language. In this work, our primary goal is to bridge the language gap
by building a human-curated instruction-following dataset spanning 65
languages. We worked with fluent speakers of languages from around the world to
collect natural instances of instructions and completions. Furthermore, we
create the most extensive multilingual collection to date, comprising 513
million instances through templating and translating existing datasets across
114 languages. In total, we contribute four key resources: we develop and
open-source the Aya Annotation Platform, the Aya Dataset, the Aya Collection,
and the Aya Evaluation Suite. The Aya initiative also serves as a valuable case
study in participatory research, involving collaborators from 119 countries. We
see this as a valuable framework for future research collaborations that aim to
bridge gaps in resources. | cs.CL | Aya Dataset: An Open-Access Collection for Multilingual Instruction Tuning |
2024-02-09T18:50:51Z | http://arxiv.org/pdf/2402.06617v1 | http://arxiv.org/abs/2402.06617v1 | Mostafa Masumi, Seyed Soroush Majd, Mehrnoush Shamsfard, Hamid Beigy | 2024-02-09T18:50:51Z | cs.CL | We introduce FaBERT, a Persian BERT-base model pre-trained on the HmBlogs
corpus, encompassing both informal and formal Persian texts. FaBERT is designed
to excel in traditional Natural Language Understanding (NLU) tasks, addressing
the intricacies of diverse sentence structures and linguistic styles prevalent
in the Persian language. In our comprehensive evaluation of FaBERT on 12
datasets in various downstream tasks, encompassing Sentiment Analysis (SA),
Named Entity Recognition (NER), Natural Language Inference (NLI), Question
Answering (QA), and Question Paraphrasing (QP), it consistently demonstrated
improved performance, all achieved within a compact model size. The findings
highlight the importance of utilizing diverse and cleaned corpora, such as
HmBlogs, to enhance the performance of language models like BERT in Persian
Natural Language Processing (NLP) applications. FaBERT is openly accessible at
https://huggingface.co/sbunlp/fabert | cs.CL | FaBERT: Pre-training BERT on Persian Blogs |
2024-02-09T18:39:13Z | http://arxiv.org/pdf/2402.06608v2 | http://arxiv.org/abs/2402.06608v2 | Sudhir Agarwal, Anu Sreepathy | 2024-06-29T00:30:04Z | cs.CL, cs.AI | We study the problem of generating plans for given natural language planning
task requests. On one hand, LLMs excel at natural language processing but do
not perform well on planning. On the other hand, classical planning tools excel
at planning tasks but require input in a structured language such as the
Planning Domain Definition Language (PDDL). We leverage the strengths of both
the techniques by using an LLM for generating the PDDL representation (task
PDDL) of planning task requests followed by using a classical planner for
computing a plan. Unlike previous approaches that use LLMs for generating task
PDDLs directly, our approach comprises of (a) translate: using an LLM only for
generating a logically interpretable intermediate representation of natural
language task description, (b) infer: deriving additional logically dependent
information from the intermediate representation using a logic reasoner
(currently, Answer Set Programming solver), and (c) compile: generating the
target task PDDL from the base and inferred information. We observe that using
an LLM to only output the intermediate representation significantly reduces LLM
errors. Consequently, TIC approach achieves, for at least one LLM, high
accuracy on task PDDL generation for all seven domains of our evaluation
dataset. | cs.CL | TIC: Translate-Infer-Compile for accurate "text to plan" using LLMs and Logical Representations |
2024-02-09T18:12:11Z | http://arxiv.org/pdf/2402.06592v1 | http://arxiv.org/abs/2402.06592v1 | Konstantin Kolokolov, Pavel Pekichev, Karthik Raghunathan | 2024-02-09T18:12:11Z | cs.CL, cs.SD, eess.AS | We propose a novel neural network architecture based on conformer transducer
that adds contextual information flow to the ASR systems. Our method improves
the accuracy of recognizing uncommon words while not harming the word error
rate of regular words. We explore the uncommon words accuracy improvement when
we use the new model and/or shallow fusion with context language model. We
found that combination of both provides cumulative gain in uncommon words
recognition accuracy. | cs.CL | Self-consistent context aware conformer transducer for speech recognition |
2024-02-09T18:05:03Z | http://arxiv.org/pdf/2402.06584v1 | http://arxiv.org/abs/2402.06584v1 | Ehsan Latif, Gyeong-Geon Lee, Knut Neuman, Tamara Kastorff, Xiaoming Zhai | 2024-02-09T18:05:03Z | cs.CL, cs.AI | The advancement of natural language processing has paved the way for
automated scoring systems in various languages, such as German (e.g., German
BERT [G-BERT]). Automatically scoring written responses to science questions in
German is a complex task and challenging for standard G-BERT as they lack
contextual knowledge in the science domain and may be unaligned with student
writing styles. This paper developed a contextualized German Science Education
BERT (G-SciEdBERT), an innovative large language model tailored for scoring
German-written responses to science tasks. Using G-BERT, we pre-trained
G-SciEdBERT on a corpus of 50K German written science responses with 5M tokens
to the Programme for International Student Assessment (PISA) 2015. We
fine-tuned G-SciEdBERT on 59 assessment items and examined the scoring
accuracy. We then compared its performance with G-BERT. Our findings reveal a
substantial improvement in scoring accuracy with G-SciEdBERT, demonstrating a
10% increase of quadratic weighted kappa compared to G-BERT (mean accuracy
difference = 0.096, SD = 0.024). These insights underline the significance of
specialized language models like G-SciEdBERT, which is trained to enhance the
accuracy of automated scoring, offering a substantial contribution to the field
of AI in education. | cs.CL | G-SciEdBERT: A Contextualized LLM for Science Assessment Tasks in German |
2024-02-09T17:27:35Z | http://arxiv.org/pdf/2402.06563v1 | http://arxiv.org/abs/2402.06563v1 | Neslihan Suzen, Evgeny M. Mirkes, Damian Roland, Jeremy Levesley, Alexander N. Gorban, Tim J. Coats | 2024-02-09T17:27:35Z | cs.LG, cs.AI, cs.CL, cs.HC, cs.IT, math.IT | Electronic patient records (EPRs) produce a wealth of data but contain
significant missing information. Understanding and handling this missing data
is an important part of clinical data analysis and if left unaddressed could
result in bias in analysis and distortion in critical conclusions. Missing data
may be linked to health care professional practice patterns and imputation of
missing data can increase the validity of clinical decisions. This study
focuses on statistical approaches for understanding and interpreting the
missing data and machine learning based clinical data imputation using a single
centre's paediatric emergency data and the data from UK's largest clinical
audit for traumatic injury database (TARN). In the study of 56,961 data points
related to initial vital signs and observations taken on children presenting to
an Emergency Department, we have shown that missing data are likely to be
non-random and how these are linked to health care professional practice
patterns. We have then examined 79 TARN fields with missing values for 5,791
trauma cases. Singular Value Decomposition (SVD) and k-Nearest Neighbour (kNN)
based missing data imputation methods are used and imputation results against
the original dataset are compared and statistically tested. We have concluded
that the 1NN imputer is the best imputation which indicates a usual pattern of
clinical decision making: find the most similar patients and take their
attributes as imputation. | cs.LG | What is Hiding in Medicine's Dark Matter? Learning with Missing Data in Medical Practices |
2024-02-09T17:20:48Z | http://arxiv.org/pdf/2403.13812v1 | http://arxiv.org/abs/2403.13812v1 | Arslan Akram | 2024-02-09T17:20:48Z | cs.DL, cs.AI, cs.CL, cs.CY, cs.LG, stat.OT, 62P25, I.7; G.1; G.3 | Many people are interested in ChatGPT since it has become a prominent AIGC
model that provides high-quality responses in various contexts, such as
software development and maintenance. Misuse of ChatGPT might cause significant
issues, particularly in public safety and education, despite its immense
potential. The majority of researchers choose to publish their work on Arxiv.
The effectiveness and originality of future work depend on the ability to
detect AI components in such contributions. To address this need, this study
will analyze a method that can see purposely manufactured content that academic
organizations use to post on Arxiv. For this study, a dataset was created using
physics, mathematics, and computer science articles. Using the newly built
dataset, the following step is to put originality.ai through its paces. The
statistical analysis shows that Originality.ai is very accurate, with a rate of
98%. | cs.DL | Quantitative Analysis of AI-Generated Texts in Academic Research: A Study of AI Presence in Arxiv Submissions using AI Detection Tool |
2024-02-09T17:18:33Z | http://arxiv.org/pdf/2402.06559v1 | http://arxiv.org/abs/2402.06559v1 | Brian Yang, Huangyuan Su, Nikolaos Gkanatsios, Tsung-Wei Ke, Ayush Jain, Jeff Schneider, Katerina Fragkiadaki | 2024-02-09T17:18:33Z | cs.LG, cs.AI, cs.CL, cs.RO | Diffusion models excel at modeling complex and multimodal trajectory
distributions for decision-making and control. Reward-gradient guided denoising
has been recently proposed to generate trajectories that maximize both a
differentiable reward function and the likelihood under the data distribution
captured by a diffusion model. Reward-gradient guided denoising requires a
differentiable reward function fitted to both clean and noised samples,
limiting its applicability as a general trajectory optimizer. In this paper, we
propose DiffusionES, a method that combines gradient-free optimization with
trajectory denoising to optimize black-box non-differentiable objectives while
staying in the data manifold. Diffusion-ES samples trajectories during
evolutionary search from a diffusion model and scores them using a black-box
reward function. It mutates high-scoring trajectories using a truncated
diffusion process that applies a small number of noising and denoising steps,
allowing for much more efficient exploration of the solution space. We show
that DiffusionES achieves state-of-the-art performance on nuPlan, an
established closed-loop planning benchmark for autonomous driving. Diffusion-ES
outperforms existing sampling-based planners, reactive deterministic or
diffusion-based policies, and reward-gradient guidance. Additionally, we show
that unlike prior guidance methods, our method can optimize non-differentiable
language-shaped reward functions generated by few-shot LLM prompting. When
guided by a human teacher that issues instructions to follow, our method can
generate novel, highly complex behaviors, such as aggressive lane weaving,
which are not present in the training data. This allows us to solve the hardest
nuPlan scenarios which are beyond the capabilities of existing trajectory
optimization methods and driving policies. | cs.LG | Diffusion-ES: Gradient-free Planning with Diffusion for Autonomous Driving and Zero-Shot Instruction Following |
2024-02-09T17:02:41Z | http://arxiv.org/pdf/2402.06549v1 | http://arxiv.org/abs/2402.06549v1 | Marek Šuppa, Daniel Skala, Daniela Jašš, Samuel Sučík, Andrej Švec, Peter Hraška | 2024-02-09T17:02:41Z | cs.CL, cs.AI | This study details our approach for the CASE 2024 Shared Task on Climate
Activism Stance and Hate Event Detection, focusing on Hate Speech Detection,
Hate Speech Target Identification, and Stance Detection as classification
challenges. We explored the capability of Large Language Models (LLMs),
particularly GPT-4, in zero- or few-shot settings enhanced by retrieval
augmentation and re-ranking for Tweet classification. Our goal was to determine
if LLMs could match or surpass traditional methods in this context.
We conducted an ablation study with LLaMA for comparison, and our results
indicate that our models significantly outperformed the baselines, securing
second place in the Target Detection task. The code for our submission is
available at https://github.com/NaiveNeuron/bryndza-case-2024 | cs.CL | Bryndza at ClimateActivism 2024: Stance, Target and Hate Event Detection via Retrieval-Augmented GPT-4 and LLaMA |
2024-02-09T17:00:32Z | http://arxiv.org/pdf/2402.06544v1 | http://arxiv.org/abs/2402.06544v1 | Yukun Huang, Yixin Liu, Raghuveer Thirukovalluru, Arman Cohan, Bhuwan Dhingra | 2024-02-09T17:00:32Z | cs.CL, cs.AI, cs.LG | To enhance Large Language Models' (LLMs) reliability, calibration is
essential -- the model's assessed confidence scores should align with the
actual likelihood of its responses being correct. However, current confidence
elicitation methods and calibration metrics typically rely on a binary
true/false assessment of response correctness. This approach does not apply to
long-form generation, where an answer can be partially correct. Addressing this
gap, we introduce a unified calibration framework, in which both the
correctness of the LLMs' responses and their associated confidence levels are
treated as distributions across a range of scores. Within this framework, we
develop three metrics to precisely evaluate LLM calibration and further propose
two confidence elicitation methods based on self-consistency and
self-evaluation. Our experiments, which include long-form QA and summarization
tasks, demonstrate that larger models don't necessarily guarantee better
calibration, that calibration performance is found to be metric-dependent, and
that self-consistency methods excel in factoid datasets. We also find that
calibration can be enhanced through techniques such as fine-tuning, integrating
relevant source documents, scaling the temperature, and combining
self-consistency with self-evaluation. Lastly, we showcase a practical
application of our system: selecting and cascading open-source models and
ChatGPT to optimize correctness given a limited API budget. This research not
only challenges existing notions of LLM calibration but also offers practical
methodologies for improving trustworthiness in long-form generation. | cs.CL | Calibrating Long-form Generations from Large Language Models |
2024-02-09T16:53:19Z | http://arxiv.org/pdf/2402.10231v1 | http://arxiv.org/abs/2402.10231v1 | Naveed Ejaz, Fakhra Kashif, Salimur Choudhury | 2024-02-09T16:53:19Z | cs.SI, cs.CL, cs.LG, I.2 | In recent years, the rising use of social media has propelled automated
cyberbullying detection into a prominent research domain. However, challenges
persist due to the absence of a standardized definition and universally
accepted datasets. Many researchers now view cyberbullying as a facet of
cyberaggression, encompassing factors like repetition, peer relationships, and
harmful intent in addition to online aggression. Acquiring comprehensive data
reflective of all cyberbullying components from social media networks proves to
be a complex task. This paper provides a description of an extensive
semi-synthetic cyberbullying dataset that incorporates all of the essential
aspects of cyberbullying, including aggression, repetition, peer relationships,
and intent to harm. The method of creating the dataset is succinctly outlined,
and a detailed overview of the publicly accessible dataset is additionally
presented. This accompanying data article provides an in-depth look at the
dataset, increasing transparency and enabling replication. It also aids in a
deeper understanding of the data, supporting broader research use. | cs.SI | A Multi-faceted Semi-Synthetic Dataset for Automated Cyberbullying Detection |
2024-02-09T16:40:59Z | http://arxiv.org/pdf/2402.06529v3 | http://arxiv.org/abs/2402.06529v3 | Kaiqu Liang, Zixu Zhang, Jaime Fernández Fisac | 2024-06-04T02:25:30Z | cs.AI, cs.CL, cs.LG | Large language models (LLMs) exhibit advanced reasoning skills, enabling
robots to comprehend natural language instructions and strategically plan
high-level actions through proper grounding. However, LLM hallucination may
result in robots confidently executing plans that are misaligned with user
goals or, in extreme cases, unsafe. Additionally, inherent ambiguity in natural
language instructions can induce task uncertainty, particularly in situations
where multiple valid options exist. To address this issue, LLMs must identify
such uncertainty and proactively seek clarification. This paper explores the
concept of introspective planning as a systematic method for guiding LLMs in
forming uncertainty--aware plans for robotic task execution without the need
for fine-tuning. We investigate uncertainty quantification in task-level robot
planning and demonstrate that introspection significantly improves both success
rates and safety compared to state-of-the-art LLM-based planning approaches.
Furthermore, we assess the effectiveness of introspective planning in
conjunction with conformal prediction, revealing that this combination yields
tighter confidence bounds, thereby maintaining statistical success guarantees
with fewer superfluous user clarification queries. Code is available at
https://github.com/kevinliang888/IntroPlan. | cs.AI | Introspective Planning: Aligning Robots' Uncertainty with Inherent Task Ambiguity |
2024-02-09T16:18:38Z | http://arxiv.org/pdf/2402.06512v3 | http://arxiv.org/abs/2402.06512v3 | Wenhao Zheng, Dongsheng Peng, Hongxia Xu, Yun Li, Hongtu Zhu, Tianfan Fu, Huaxiu Yao | 2024-05-09T01:22:35Z | cs.LG, cs.CL | The clinical trial is a pivotal and costly process, often spanning multiple
years and requiring substantial financial resources. Therefore, the development
of clinical trial outcome prediction models aims to exclude drugs likely to
fail and holds the potential for significant cost savings. Recent data-driven
attempts leverage deep learning methods to integrate multimodal data for
predicting clinical trial outcomes. However, these approaches rely on manually
designed modal-specific encoders, which limits both the extensibility to adapt
new modalities and the ability to discern similar information patterns across
different modalities. To address these issues, we propose a multimodal
mixture-of-experts (LIFTED) approach for clinical trial outcome prediction.
Specifically, LIFTED unifies different modality data by transforming them into
natural language descriptions. Then, LIFTED constructs unified noise-resilient
encoders to extract information from modal-specific language descriptions.
Subsequently, a sparse Mixture-of-Experts framework is employed to further
refine the representations, enabling LIFTED to identify similar information
patterns across different modalities and extract more consistent
representations from those patterns using the same expert model. Finally, a
mixture-of-experts module is further employed to dynamically integrate
different modality representations for prediction, which gives LIFTED the
ability to automatically weigh different modalities and pay more attention to
critical information. The experiments demonstrate that LIFTED significantly
enhances performance in predicting clinical trial outcomes across all three
phases compared to the best baseline, showcasing the effectiveness of our
proposed key components. | cs.LG | Multimodal Clinical Trial Outcome Prediction with Large Language Models |
2024-02-09T16:15:30Z | http://arxiv.org/pdf/2402.06509v1 | http://arxiv.org/abs/2402.06509v1 | Alberto Testoni, Raquel Fernández | 2024-02-09T16:15:30Z | cs.CL, cs.AI | Clarification questions are an essential dialogue tool to signal
misunderstanding, ambiguities, and under-specification in language use. While
humans are able to resolve uncertainty by asking questions since childhood,
modern dialogue systems struggle to generate effective questions. To make
progress in this direction, in this work we take a collaborative dialogue task
as a testbed and study how model uncertainty relates to human uncertainty -- an
as yet under-explored problem. We show that model uncertainty does not mirror
human clarification-seeking behavior, which suggests that using human
clarification questions as supervision for deciding when to ask may not be the
most effective way to resolve model uncertainty. To address this issue, we
propose an approach to generating clarification questions based on model
uncertainty estimation, compare it to several alternatives, and show that it
leads to significant improvements in terms of task success. Our findings
highlight the importance of equipping dialogue systems with the ability to
assess their own uncertainty and exploit in interaction. | cs.CL | Asking the Right Question at the Right Time: Human and Model Uncertainty Guidance to Ask Clarification Questions |
2024-02-09T16:11:04Z | http://arxiv.org/pdf/2402.06501v2 | http://arxiv.org/abs/2402.06501v2 | Anna Madison, Ellen Novoseller, Vinicius G. Goecks, Benjamin T. Files, Nicholas Waytowich, Alfred Yu, Vernon J. Lawhern, Steven Thurman, Christopher Kelshaw, Kaleb McDowell | 2024-03-28T15:17:01Z | cs.LG, cs.AI, cs.CL, cs.HC, I.2.6; I.2.7; J.7 | Future warfare will require Command and Control (C2) personnel to make
decisions at shrinking timescales in complex and potentially ill-defined
situations. Given the need for robust decision-making processes and
decision-support tools, integration of artificial and human intelligence holds
the potential to revolutionize the C2 operations process to ensure adaptability
and efficiency in rapidly changing operational environments. We propose to
leverage recent promising breakthroughs in interactive machine learning, in
which humans can cooperate with machine learning algorithms to guide machine
learning algorithm behavior. This paper identifies several gaps in
state-of-the-art science and technology that future work should address to
extend these approaches to function in complex C2 contexts. In particular, we
describe three research focus areas that together, aim to enable scalable
interactive machine learning (SIML): 1) developing human-AI interaction
algorithms to enable planning in complex, dynamic situations; 2) fostering
resilient human-AI teams through optimizing roles, configurations, and trust;
and 3) scaling algorithms and human-AI teams for flexibility across a range of
potential contexts and situations. | cs.LG | Scalable Interactive Machine Learning for Future Command and Control |
2024-02-09T16:10:29Z | http://arxiv.org/pdf/2402.07946v2 | http://arxiv.org/abs/2402.07946v2 | Kaleb McDowell, Ellen Novoseller, Anna Madison, Vinicius G. Goecks, Christopher Kelshaw | 2024-03-28T15:17:30Z | cs.HC, cs.AI, cs.CL, cs.LG, I.2.6; I.2.7; J.7 | Future warfare will require Command and Control (C2) decision-making to occur
in more complex, fast-paced, ill-structured, and demanding conditions. C2 will
be further complicated by operational challenges such as Denied, Degraded,
Intermittent, and Limited (DDIL) communications and the need to account for
many data streams, potentially across multiple domains of operation. Yet,
current C2 practices -- which stem from the industrial era rather than the
emerging intelligence era -- are linear and time-consuming. Critically, these
approaches may fail to maintain overmatch against adversaries on the future
battlefield. To address these challenges, we propose a vision for future C2
based on robust partnerships between humans and artificial intelligence (AI)
systems. This future vision is encapsulated in three operational impacts:
streamlining the C2 operations process, maintaining unity of effort, and
developing adaptive collective knowledge systems. This paper illustrates the
envisaged future C2 capabilities, discusses the assumptions that shaped them,
and describes how the proposed developments could transform C2 in future
warfare. | cs.HC | Re-Envisioning Command and Control |
2024-02-09T15:53:15Z | http://arxiv.org/pdf/2402.06492v1 | http://arxiv.org/abs/2402.06492v1 | Yichen Jiang, Xiang Zhou, Mohit Bansal | 2024-02-09T15:53:15Z | cs.CL, cs.AI, cs.LG | Transformers generalize to novel compositions of structures and entities
after being trained on a complex dataset, but easily overfit on datasets of
insufficient complexity. We observe that when the training set is sufficiently
complex, the model encodes sentences that have a common syntactic structure
using a systematic attention pattern. Inspired by this observation, we propose
SQ-Transformer (Structurally Quantized) that explicitly encourages
systematicity in the embeddings and attention layers, even with a training set
of low complexity. At the embedding level, we introduce Structure-oriented
Vector Quantization (SoVQ) to cluster word embeddings into several classes of
structurally equivalent entities. At the attention level, we devise the
Systematic Attention Layer (SAL) and an alternative, Systematically Regularized
Layer (SRL) that operate on the quantized word embeddings so that sentences of
the same structure are encoded with invariant or similar attention patterns.
Empirically, we show that SQ-Transformer achieves stronger compositional
generalization than the vanilla Transformer on multiple low-complexity semantic
parsing and machine translation datasets. In our analysis, we show that SoVQ
indeed learns a syntactically clustered embedding space and SAL/SRL induces
generalizable attention patterns, which lead to improved systematicity. | cs.CL | Inducing Systematicity in Transformers by Attending to Structurally Quantized Embeddings |
2024-02-09T15:02:56Z | http://arxiv.org/pdf/2402.06457v1 | http://arxiv.org/abs/2402.06457v1 | Arian Hosseini, Xingdi Yuan, Nikolay Malkin, Aaron Courville, Alessandro Sordoni, Rishabh Agarwal | 2024-02-09T15:02:56Z | cs.LG, cs.AI, cs.CL | Common self-improvement approaches for large language models (LLMs), such as
STaR (Zelikman et al., 2022), iteratively fine-tune LLMs on self-generated
solutions to improve their problem-solving ability. However, these approaches
discard the large amounts of incorrect solutions generated during this process,
potentially neglecting valuable information in such solutions. To address this
shortcoming, we propose V-STaR that utilizes both the correct and incorrect
solutions generated during the self-improvement process to train a verifier
using DPO that judges correctness of model-generated solutions. This verifier
is used at inference time to select one solution among many candidate
solutions. Running V-STaR for multiple iterations results in progressively
better reasoners and verifiers, delivering a 4% to 17% test accuracy
improvement over existing self-improvement and verification approaches on
common code generation and math reasoning benchmarks with LLaMA2 models. | cs.LG | V-STaR: Training Verifiers for Self-Taught Reasoners |
2024-02-09T14:39:20Z | http://arxiv.org/pdf/2402.06443v1 | http://arxiv.org/abs/2402.06443v1 | Recep Firat Cekinel, Pinar Karagoz | 2024-02-09T14:39:20Z | cs.CL | The rapid dissemination of misinformation through social media increased the
importance of automated fact-checking. Furthermore, studies on what deep neural
models pay attention to when making predictions have increased in recent years.
While significant progress has been made in this field, it has not yet reached
a level of reasoning comparable to human reasoning. To address these gaps, we
propose a multi-task explainable neural model for misinformation detection.
Specifically, this work formulates an explanation generation process of the
model's veracity prediction as a text summarization problem. Additionally, the
performance of the proposed model is discussed on publicly available datasets
and the findings are evaluated with related studies. | cs.CL | Explaining Veracity Predictions with Evidence Summarization: A Multi-Task Model Approach |
2024-02-09T14:08:23Z | http://arxiv.org/pdf/2402.06420v1 | http://arxiv.org/abs/2402.06420v1 | Yvette Graham, Mohammed Rameez Qureshi, Haider Khalid, Gerasimos Lampouras, Ignacio Iacobacci, Qun Liu | 2024-02-09T14:08:23Z | cs.CL | The aim of this workshop is to bring together experts working on open-domain
dialogue research. In this speedily advancing research area many challenges
still exist, such as learning information from conversations, engaging in
realistic and convincing simulation of human intelligence and reasoning.
SCI-CHAT follows previous workshops on open domain dialogue but with a focus on
the simulation of intelligent conversation as judged in a live human
evaluation. Models aim to include the ability to follow a challenging topic
over a multi-turn conversation, while positing, refuting and reasoning over
arguments. The workshop included both a research track and shared task. The
main goal of this paper is to provide an overview of the shared task and a link
to an additional paper that will include an in depth analysis of the shared
task results following presentation at the workshop. | cs.CL | Findings of the First Workshop on Simulating Conversational Intelligence in Chat |
2024-02-09T12:10:00Z | http://arxiv.org/pdf/2402.06360v1 | http://arxiv.org/abs/2402.06360v1 | Peiyuan Gong, Jiamian Li, Jiaxin Mao | 2024-02-09T12:10:00Z | cs.IR, cs.AI, cs.CL | Collaborative search supports multiple users working together to accomplish a
specific search task. Research has found that designing lightweight
collaborative search plugins within instant messaging platforms aligns better
with users' collaborative habits. However, due to the complexity of multi-user
interaction scenarios, it is challenging to implement a fully functioning
lightweight collaborative search system. Therefore, previous studies on
lightweight collaborative search had to rely on the Wizard of Oz paradigm. In
recent years, large language models (LLMs) have been demonstrated to interact
naturally with users and achieve complex information-seeking tasks through
LLM-based agents. Hence, to better support the research in collaborative
search, in this demo, we propose CoSearchAgent, a lightweight collaborative
search agent powered by LLMs. CoSearchAgent is designed as a Slack plugin that
can support collaborative search during multi-party conversations on this
platform. Equipped with the capacity to understand the queries and context in
multi-user conversations and the ability to search the Web for relevant
information via APIs, CoSearchAgent can respond to user queries with answers
grounded on the relevant search results. It can also ask clarifying questions
when the information needs are unclear. The proposed CoSearchAgent is highly
flexible and would be useful for supporting further research on collaborative
search. The code and demo video are accessible. | cs.IR | CoSearchAgent: A Lightweight Collaborative Search Agent with Large Language Models |
2024-02-09T11:34:39Z | http://arxiv.org/pdf/2402.06342v1 | http://arxiv.org/abs/2402.06342v1 | Harritxu Gete, Thierry Etchegoyhen | 2024-02-09T11:34:39Z | cs.CL | Standard context-aware neural machine translation (NMT) typically relies on
parallel document-level data, exploiting both source and target contexts.
Concatenation-based approaches in particular, still a strong baseline for
document-level NMT, prepend source and/or target context sentences to the
sentences to be translated, with model variants that exploit equal amounts of
source and target data on each side achieving state-of-the-art results. In this
work, we investigate whether target data should be further promoted within
standard concatenation-based approaches, as most document-level phenomena rely
on information that is present on the target language side. We evaluate novel
concatenation-based variants where the target context is prepended to the
source language, either in isolation or in combination with the source context.
Experimental results in English-Russian and Basque-Spanish show that including
target context in the source leads to large improvements on target language
phenomena. On source-dependent phenomena, using only target language context in
the source achieves parity with state-of-the-art concatenation approaches, or
slightly underperforms, whereas combining source and target context on the
source side leads to significant gains across the board. | cs.CL | Promoting Target Data in Context-aware Neural Machine Translation |
2024-02-09T11:34:16Z | http://arxiv.org/pdf/2402.06341v1 | http://arxiv.org/abs/2402.06341v1 | Xuanzhong Chen, Xiaohao Mao, Qihan Guo, Lun Wang, Shuyang Zhang, Ting Chen | 2024-02-09T11:34:16Z | cs.CL | Generalist Large Language Models (LLMs), such as GPT-4, have shown
considerable promise in various domains, including medical diagnosis. Rare
diseases, affecting approximately 300 million people worldwide, often have
unsatisfactory clinical diagnosis rates primarily due to a lack of experienced
physicians and the complexity of differentiating among many rare diseases. In
this context, recent news such as "ChatGPT correctly diagnosed a 4-year-old's
rare disease after 17 doctors failed" underscore LLMs' potential, yet
underexplored, role in clinically diagnosing rare diseases. To bridge this
research gap, we introduce RareBench, a pioneering benchmark designed to
systematically evaluate the capabilities of LLMs on 4 critical dimensions
within the realm of rare diseases. Meanwhile, we have compiled the largest
open-source dataset on rare disease patients, establishing a benchmark for
future studies in this domain. To facilitate differential diagnosis of rare
diseases, we develop a dynamic few-shot prompt methodology, leveraging a
comprehensive rare disease knowledge graph synthesized from multiple knowledge
bases, significantly enhancing LLMs' diagnostic performance. Moreover, we
present an exhaustive comparative study of GPT-4's diagnostic capabilities
against those of specialist physicians. Our experimental findings underscore
the promising potential of integrating LLMs into the clinical diagnostic
process for rare diseases. This paves the way for exciting possibilities in
future advancements in this field. | cs.CL | RareBench: Can LLMs Serve as Rare Diseases Specialists? |
2024-02-09T11:23:14Z | http://arxiv.org/pdf/2402.06334v1 | http://arxiv.org/abs/2402.06334v1 | Fernando Ferraretto, Thiago Laitz, Roberto Lotufo, Rodrigo Nogueira | 2024-02-09T11:23:14Z | cs.IR, cs.AI, cs.CL | ExaRanker recently introduced an approach to training information retrieval
(IR) models, incorporating natural language explanations as additional labels.
The method addresses the challenge of limited labeled examples, leading to
improvements in the effectiveness of IR models. However, the initial results
were based on proprietary language models such as GPT-3.5, which posed
constraints on dataset size due to its cost and data privacy. In this paper, we
introduce ExaRanker-Open, where we adapt and explore the use of open-source
language models to generate explanations. The method has been tested using
different LLMs and datasets sizes to better comprehend the effective
contribution of data augmentation. Our findings reveal that incorporating
explanations consistently enhances neural rankers, with benefits escalating as
the LLM size increases. Notably, the data augmentation method proves
advantageous even with large datasets, as evidenced by ExaRanker surpassing the
target baseline by 0.6 nDCG@10 points in our study. To encourage further
advancements by the research community, we have open-sourced both the code and
datasets at https://github.com/unicamp-dl/ExaRanker. | cs.IR | ExaRanker-Open: Synthetic Explanation for IR using Open-Source LLMs |
2024-02-09T11:22:08Z | http://arxiv.org/pdf/2402.06332v2 | http://arxiv.org/abs/2402.06332v2 | Huaiyuan Ying, Shuo Zhang, Linyang Li, Zhejian Zhou, Yunfan Shao, Zhaoye Fei, Yichuan Ma, Jiawei Hong, Kuikun Liu, Ziyi Wang, Yudong Wang, Zijian Wu, Shuaibin Li, Fengzhe Zhou, Hongwei Liu, Songyang Zhang, Wenwei Zhang, Hang Yan, Xipeng Qiu, Jiayu Wang, Kai Chen, Dahua Lin | 2024-05-24T07:09:21Z | cs.CL | The math abilities of large language models can represent their abstract
reasoning ability. In this paper, we introduce and open-source our math
reasoning LLMs InternLM-Math which is continue pre-trained from InternLM2. We
unify chain-of-thought reasoning, reward modeling, formal reasoning, data
augmentation, and code interpreter in a unified seq2seq format and supervise
our model to be a versatile math reasoner, verifier, prover, and augmenter.
These abilities can be used to develop the next math LLMs or self-iteration.
InternLM-Math obtains open-sourced state-of-the-art performance under the
setting of in-context learning, supervised fine-tuning, and code-assisted
reasoning in various informal and formal benchmarks including GSM8K, MATH,
Hungary math exam, MathBench-ZH, and MiniF2F. Our pre-trained model achieves
30.3 on the MiniF2F test set without fine-tuning. We further explore how to use
LEAN to solve math problems and study its performance under the setting of
multi-task learning which shows the possibility of using LEAN as a unified
platform for solving and proving in math. Our models, codes, and data are
released at \url{https://github.com/InternLM/InternLM-Math}. | cs.CL | InternLM-Math: Open Math Large Language Models Toward Verifiable Reasoning |
2024-02-09T10:25:01Z | http://arxiv.org/pdf/2402.18589v1 | http://arxiv.org/abs/2402.18589v1 | Miloš Košprdić, Adela Ljajić, Bojana Bašaragin, Darija Medvecki, Nikola Milošević | 2024-02-09T10:25:01Z | cs.IR, cs.AI, cs.CL, cs.LG | In this paper, we present the current progress of the project Verif.ai, an
open-source scientific generative question-answering system with referenced and
verified answers. The components of the system are (1) an information retrieval
system combining semantic and lexical search techniques over scientific papers
(PubMed), (2) a fine-tuned generative model (Mistral 7B) taking top answers and
generating answers with references to the papers from which the claim was
derived, and (3) a verification engine that cross-checks the generated claim
and the abstract or paper from which the claim was derived, verifying whether
there may have been any hallucinations in generating the claim. We are
reinforcing the generative model by providing the abstract in context, but in
addition, an independent set of methods and models are verifying the answer and
checking for hallucinations. Therefore, we believe that by using our method, we
can make scientists more productive, while building trust in the use of
generative language models in scientific environments, where hallucinations and
misinformation cannot be tolerated. | cs.IR | Verif.ai: Towards an Open-Source Scientific Generative Question-Answering System with Referenced and Verifiable Answers |
2024-02-09T09:44:06Z | http://arxiv.org/pdf/2402.10948v2 | http://arxiv.org/abs/2402.10948v2 | Wenyu Li, Yinuo Zhu, Xin Lin, Ming Li, Ziyue Jiang, Ziqian Zeng | 2024-03-15T02:02:02Z | cs.CL, cs.AI | Traditional discriminative approaches in mental health analysis are known for
their strong capacity but lack interpretability and demand large-scale
annotated data. The generative approaches, such as those based on large
language models (LLMs), have the potential to get rid of heavy annotations and
provide explanations but their capabilities still fall short compared to
discriminative approaches, and their explanations may be unreliable due to the
fact that the generation of explanation is a black-box process. Inspired by the
psychological assessment practice of using scales to evaluate mental states,
our method which is called Mental Analysis by Incorporating Mental Scales
(MAIMS), incorporates two procedures via LLMs. First, the patient completes
mental scales, and second, the psychologist interprets the collected
information from the mental scales and makes informed decisions. Experimental
results show that MAIMS outperforms other zero-shot methods. MAIMS can generate
more rigorous explanation based on the outputs of mental scales | cs.CL | Zero-shot Explainable Mental Health Analysis on Social Media by Incorporating Mental Scales |
2024-02-09T09:25:18Z | http://arxiv.org/pdf/2402.06264v1 | http://arxiv.org/abs/2402.06264v1 | Unggi Lee, Minji Jeon, Yunseo Lee, Gyuri Byun, Yoorim Son, Jaeyoon Shin, Hongkyu Ko, Hyeoncheol Kim | 2024-02-09T09:25:18Z | cs.AI, cs.CL, cs.SI | Art appreciation is vital in nurturing critical thinking and emotional
intelligence among learners. However, traditional art appreciation education
has often been hindered by limited access to art resources, especially for
disadvantaged students, and an imbalanced emphasis on STEM subjects in
mainstream education. In response to these challenges, recent technological
advancements have paved the way for innovative solutions. This study explores
the application of multi-modal large language models (MLLMs) in art
appreciation education, focusing on developing LLaVA-Docent, a model that
leverages these advancements. Our approach involved a comprehensive literature
review and consultations with experts in the field, leading to developing a
robust data framework. Utilizing this framework, we generated a virtual
dialogue dataset that was leveraged by GPT-4. This dataset was instrumental in
training the MLLM, named LLaVA-Docent. Six researchers conducted quantitative
and qualitative evaluations of LLaVA-Docent to assess its effectiveness,
benchmarking it against the GPT-4 model in a few-shot setting. The evaluation
process revealed distinct strengths and weaknesses of the LLaVA-Docent model.
Our findings highlight the efficacy of LLaVA-Docent in enhancing the
accessibility and engagement of art appreciation education. By harnessing the
potential of MLLMs, this study makes a significant contribution to the field of
art education, proposing a novel methodology that reimagines the way art
appreciation is taught and experienced. | cs.AI | LLaVA-Docent: Instruction Tuning with Multimodal Large Language Model to Support Art Appreciation Education |
2024-02-09T09:20:59Z | http://arxiv.org/pdf/2402.06262v2 | http://arxiv.org/abs/2402.06262v2 | Siyu Ren, Kenny Q. Zhu | 2024-02-17T10:08:14Z | cs.CL, cs.AI | Despite the recent success associated with Large Language Models (LLMs), they
are notably cost-prohibitive to deploy in resource-constrained environments due
to their excessive memory and computational demands. In addition to model
parameters, the key-value cache is also stored in GPU memory, growing linearly
with batch size and sequence length. As a remedy, recent works have proposed
various eviction policies for maintaining the overhead of key-value cache under
a given budget. This paper embarks on the efficacy of existing eviction
policies in terms of importance score calculation and eviction scope
construction. We identify the deficiency of prior policies in these two aspects
and introduce RoCo, a robust cache omission policy based on temporal attention
scores and robustness measures. Extensive experimentation spanning prefilling
and auto-regressive decoding stages validates the superiority of RoCo. Finally,
we release EasyKV, a versatile software package dedicated to user-friendly
key-value constrained generative inference. Code available at
https://github.com/DRSY/EasyKV. | cs.CL | On the Efficacy of Eviction Policy for Key-Value Constrained Generative Language Model Inference |
2024-02-09T09:09:39Z | http://arxiv.org/pdf/2402.06255v2 | http://arxiv.org/abs/2402.06255v2 | Yichuan Mo, Yuji Wang, Zeming Wei, Yisen Wang | 2024-06-09T16:18:46Z | cs.LG, cs.AI, cs.CL, cs.CR | While Large Language Models (LLMs) have achieved tremendous success in
various applications, they are also susceptible to jailbreak attacks. Several
primary defense strategies have been proposed to protect LLMs from producing
harmful information, mostly with a particular focus on harmful content
filtering or heuristical defensive prompt designs. However, how to achieve
intrinsic robustness through the prompts remains an open problem. In this
paper, motivated by adversarial training paradigms for achieving reliable
robustness, we propose an approach named Prompt Adversarial Tuning (PAT) that
trains a prompt control attached to the user prompt as a guard prefix. To
achieve our defense goal whilst maintaining natural performance, we optimize
the control prompt with both adversarial and benign prompts. Comprehensive
experiments show that our method is effective against both black-box and
white-box attacks, reducing the success rate of advanced attacks to nearly 0
while maintaining the model's utility on the benign task. The proposed defense
strategy incurs only negligible computational overhead, charting a new
perspective for future explorations in LLM security. Our code is available at
https://github.com/rain152/PAT. | cs.LG | Fight Back Against Jailbreaking via Prompt Adversarial Tuning |
2024-02-09T07:13:44Z | http://arxiv.org/pdf/2402.06221v2 | http://arxiv.org/abs/2402.06221v2 | Saurabh Bhausaheb Zinjad, Amrita Bhattacharjee, Amey Bhilegaonkar, Huan Liu | 2024-05-08T03:09:10Z | cs.CL, cs.IR | Crafting the ideal, job-specific resume is a challenging task for many job
applicants, especially for early-career applicants. While it is highly
recommended that applicants tailor their resume to the specific role they are
applying for, manually tailoring resumes to job descriptions and role-specific
requirements is often (1) extremely time-consuming, and (2) prone to human
errors. Furthermore, performing such a tailoring step at scale while applying
to several roles may result in a lack of quality of the edited resumes. To
tackle this problem, in this demo paper, we propose ResumeFlow: a Large
Language Model (LLM) aided tool that enables an end user to simply provide
their detailed resume and the desired job posting, and obtain a personalized
resume specifically tailored to that specific job posting in the matter of a
few seconds. Our proposed pipeline leverages the language understanding and
information extraction capabilities of state-of-the-art LLMs such as OpenAI's
GPT-4 and Google's Gemini, in order to (1) extract details from a job
description, (2) extract role-specific details from the user-provided resume,
and then (3) use these to refine and generate a role-specific resume for the
user. Our easy-to-use tool leverages the user-chosen LLM in a completely
off-the-shelf manner, thus requiring no fine-tuning. We demonstrate the
effectiveness of our tool via a video demo and propose novel task-specific
evaluation metrics to control for alignment and hallucination. Our tool is
available at https://job-aligned-resume.streamlit.app. | cs.CL | ResumeFlow: An LLM-facilitated Pipeline for Personalized Resume Generation and Refinement |
2024-02-09T07:12:56Z | http://arxiv.org/pdf/2402.06220v1 | http://arxiv.org/abs/2402.06220v1 | Lu Chen, Wei Huang, Ruqing Zhang, Wei Chen, Jiafeng Guo, Xueqi Cheng | 2024-02-09T07:12:56Z | cs.CL | Instruction tuning on a mixture of tasks has improved zero-shot capabilities
in natural language processing (NLP). Nevertheless, existing methods often
learn features that exhibit correlations between instruction-formatted samples
and target labels, rather than causal relationships. Termed as ``spurious
correlation'' in statistics, such a correlation may change drastically in a new
task, making the effect from the learned features to be misleading. To this
end, we develop a meta Structural Causal Model (meta-SCM) to integrate
different NLP tasks under a single causal structure of the data. Specifically,
the meta-SCM introduces multiple latent factors that represent properties of
source context, only some of which causally influence the target labels for a
specific task. The key idea is to learn task-required causal factors and only
use those to make predictions for a given task. Theoretically, we prove the
causal factor can be identified without mixing information from others. Guided
by the identifiability, we propose a Structural Instruction Tuning (SIT) method
to learn the task-required causal representations that can mimic the causal
factors for each task. The utility of our approach is verified by improvements
of zero-shot ability on a range of unseen datasets and tasks. | cs.CL | A Unified Causal View of Instruction Tuning |
2024-02-09T06:16:08Z | http://arxiv.org/pdf/2402.06204v1 | http://arxiv.org/abs/2402.06204v1 | Juhyun Oh, Eunsu Kim, Inha Cha, Alice Oh | 2024-02-09T06:16:08Z | cs.CL, cs.AI | This paper explores the assumption that Large Language Models (LLMs) skilled
in generation tasks are equally adept as evaluators. We assess the performance
of three LLMs and one open-source LM in Question-Answering (QA) and evaluation
tasks using the TriviaQA (Joshi et al., 2017) dataset. Results indicate a
significant disparity, with LLMs exhibiting lower performance in evaluation
tasks compared to generation tasks. Intriguingly, we discover instances of
unfaithful evaluation where models accurately evaluate answers in areas where
they lack competence, underscoring the need to examine the faithfulness and
trustworthiness of LLMs as evaluators. This study contributes to the
understanding of "the Generative AI Paradox" (West et al., 2023), highlighting
a need to explore the correlation between generative excellence and evaluation
proficiency, and the necessity to scrutinize the faithfulness aspect in model
evaluations. | cs.CL | The Generative AI Paradox on Evaluation: What It Can Solve, It May Not Evaluate |
2024-02-09T05:37:09Z | http://arxiv.org/pdf/2402.06196v2 | http://arxiv.org/abs/2402.06196v2 | Shervin Minaee, Tomas Mikolov, Narjes Nikzad, Meysam Chenaghlu, Richard Socher, Xavier Amatriain, Jianfeng Gao | 2024-02-20T13:33:49Z | cs.CL, cs.AI | Large Language Models (LLMs) have drawn a lot of attention due to their
strong performance on a wide range of natural language tasks, since the release
of ChatGPT in November 2022. LLMs' ability of general-purpose language
understanding and generation is acquired by training billions of model's
parameters on massive amounts of text data, as predicted by scaling laws
\cite{kaplan2020scaling,hoffmann2022training}. The research area of LLMs, while
very recent, is evolving rapidly in many different ways. In this paper, we
review some of the most prominent LLMs, including three popular LLM families
(GPT, LLaMA, PaLM), and discuss their characteristics, contributions and
limitations. We also give an overview of techniques developed to build, and
augment LLMs. We then survey popular datasets prepared for LLM training,
fine-tuning, and evaluation, review widely used LLM evaluation metrics, and
compare the performance of several popular LLMs on a set of representative
benchmarks. Finally, we conclude the paper by discussing open challenges and
future research directions. | cs.CL | Large Language Models: A Survey |
2024-02-09T04:02:43Z | http://arxiv.org/pdf/2402.10946v1 | http://arxiv.org/abs/2402.10946v1 | Cheng Li, Mengzhou Chen, Jindong Wang, Sunayana Sitaram, Xing Xie | 2024-02-09T04:02:43Z | cs.CL, cs.AI, cs.LG | Large language models (LLMs) are reported to be partial to certain cultures
owing to the training data dominance from the English corpora. Since
multilingual cultural data are often expensive to collect, existing efforts
handle this by prompt engineering or culture-specific pre-training. However,
they might overlook the knowledge deficiency of low-resource culture and
require extensive computing resources. In this paper, we propose CultureLLM, a
cost-effective solution to incorporate cultural differences into LLMs.
CultureLLM adopts World Value Survey (WVS) as seed data and generates
semantically equivalent training data via the proposed semantic data
augmentation. Using only 50 seed samples from WVS with augmented data, we
fine-tune culture-specific LLMs and one unified model (CultureLLM-One) for 9
cultures covering rich and low-resource languages. Extensive experiments on 60
culture-related datasets demonstrate that CultureLLM significantly outperforms
various counterparts such as GPT-3.5 (by 8.1%) and Gemini Pro (by 9.5%) with
comparable performance to GPT-4 or even better. Our human study shows that the
generated samples are semantically equivalent to the original samples,
providing an effective solution for LLMs augmentation. | cs.CL | CultureLLM: Incorporating Cultural Differences into Large Language Models |
2024-02-09T03:08:12Z | http://arxiv.org/pdf/2402.06155v1 | http://arxiv.org/abs/2402.06155v1 | John Hewitt, Sarah Chen, Lanruo Lora Xie, Edward Adams, Percy Liang, Christopher D. Manning | 2024-02-09T03:08:12Z | cs.CL | We introduce model editing with canonical examples, a setting in which (1) a
single learning example is provided per desired behavior, (2) evaluation is
performed exclusively out-of-distribution, and (3) deviation from an initial
model is strictly limited. A canonical example is a simple instance of good
behavior, e.g., The capital of Mauritius is Port Louis) or bad behavior, e.g.,
An aspect of researchers is coldhearted). The evaluation set contains more
complex examples of each behavior (like a paragraph in which the capital of
Mauritius is called for.) We create three datasets and modify three more for
model editing with canonical examples, covering knowledge-intensive
improvements, social bias mitigation, and syntactic edge cases. In our
experiments on Pythia language models, we find that LoRA outperforms full
finetuning and MEMIT. We then turn to the Backpack language model architecture
because it is intended to enable targeted improvement. The Backpack defines a
large bank of sense vectors--a decomposition of the different uses of each
word--which are weighted and summed to form the output logits of the model. We
propose sense finetuning, which selects and finetunes a few ($\approx$ 10)
sense vectors for each canonical example, and find that it outperforms other
finetuning methods, e.g., 4.8% improvement vs 0.3%. Finally, we improve
GPT-J-6B by an inference-time ensemble with just the changes from sense
finetuning of a 35x smaller Backpack, in one setting outperforming editing
GPT-J itself (4.1% vs 1.0%). | cs.CL | Model Editing with Canonical Examples |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.