publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-04-01T09:35:06Z | http://arxiv.org/pdf/2404.01015v1 | http://arxiv.org/abs/2404.01015v1 | ChaeHun Park, Minseok Choi, Dohyun Lee, Jaegul Choo | 2024-04-01T09:35:06Z | cs.CL | Building a reliable and automated evaluation metric is a necessary but
challenging problem for open-domain dialogue systems. Recent studies proposed
evaluation metrics that assess generated responses by considering their
relevance to previous dialogue histories. Although effective, these metrics
evaluate individual responses directly rather than considering their relative
quality compared to other responses. To handle this, we propose PairEval, a
novel dialogue evaluation metric for assessing responses by comparing their
quality against responses in different conversations. PairEval is built on top
of open-sourced and moderate-size language models, and we make them specialized
in pairwise comparison between dialogue responses. Extensive experiments on
multiple benchmarks demonstrate that our metric exhibits a higher correlation
with human judgments than baseline metrics. We also find that the proposed
comparative metric is more robust in detecting common failures from open-domain
dialogue systems, including repetition and speaker insensitivity. | cs.CL | PairEval: Open-domain Dialogue Evaluation with Pairwise Comparison |
2024-04-01T09:33:05Z | http://arxiv.org/pdf/2404.01012v2 | http://arxiv.org/abs/2404.01012v2 | Chuan Meng, Negar Arabzadeh, Arian Askari, Mohammad Aliannejadi, Maarten de Rijke | 2024-06-17T11:23:20Z | cs.IR, cs.AI, cs.CL, cs.LG, H.3.3 | Query performance prediction (QPP) aims to estimate the retrieval quality of
a search system for a query without human relevance judgments. Previous QPP
methods typically return a single scalar value and do not require the predicted
values to approximate a specific information retrieval (IR) evaluation measure,
leading to certain drawbacks: (i) a single scalar is insufficient to accurately
represent different IR evaluation measures, especially when metrics do not
highly correlate, and (ii) a single scalar limits the interpretability of QPP
methods because solely using a scalar is insufficient to explain QPP results.
To address these issues, we propose a QPP framework using automatically
generated relevance judgments (QPP-GenRE), which decomposes QPP into
independent subtasks of predicting the relevance of each item in a ranked list
to a given query. This allows us to predict any IR evaluation measure using the
generated relevance judgments as pseudo-labels. This also allows us to
interpret predicted IR evaluation measures, and identify, track and rectify
errors in generated relevance judgments to improve QPP quality. We predict an
item's relevance by using open-source large language models (LLMs) to ensure
scientific reproducibility.
We face two main challenges: (i) excessive computational costs of judging an
entire corpus for predicting a metric considering recall, and (ii) limited
performance in prompting open-source LLMs in a zero-/few-shot manner. To solve
the challenges, we devise an approximation strategy to predict an IR measure
considering recall and propose to fine-tune open-source LLMs using
human-labeled relevance judgments. Experiments on the TREC 2019-2022 deep
learning tracks show that QPP-GenRE achieves state-of-the-art QPP quality for
both lexical and neural rankers. | cs.IR | Query Performance Prediction using Relevance Judgments Generated by Large Language Models |
2024-04-01T09:24:06Z | http://arxiv.org/pdf/2404.01009v1 | http://arxiv.org/abs/2404.01009v1 | Joanito Agili Lopo, Radius Tanone | 2024-04-01T09:24:06Z | cs.CL | In Indonesia, local languages play an integral role in the culture. However,
the available Indonesian language resources still fall into the category of
limited data in the Natural Language Processing (NLP) field. This is become
problematic when build NLP model for these languages. To address this gap, we
introduce Bhinneka Korpus, a multilingual parallel corpus featuring five
Indonesian local languages. Our goal is to enhance access and utilization of
these resources, extending their reach within the country. We explained in a
detail the dataset collection process and associated challenges. Additionally,
we experimented with translation task using the IBM Model 1 due to data
constraints. The result showed that the performance of each language already
shows good indications for further development. Challenges such as lexical
variation, smoothing effects, and cross-linguistic variability are discussed.
We intend to evaluate the corpus using advanced NLP techniques for low-resource
languages, paving the way for multilingual translation models. | cs.CL | Constructing and Expanding Low-Resource and Underrepresented Parallel Datasets for Indonesian Local Languages |
2024-04-01T09:08:53Z | http://arxiv.org/pdf/2404.00999v1 | http://arxiv.org/abs/2404.00999v1 | Wei Liu, Stephen Wan, Michael Strube | 2024-04-01T09:08:53Z | cs.CL | We consider an unanswered question in the discourse processing community: why
do relation classifiers trained on explicit examples (with connectives removed)
perform poorly in real implicit scenarios? Prior work claimed this is due to
linguistic dissimilarity between explicit and implicit examples but provided no
empirical evidence. In this study, we show that one cause for such failure is a
label shift after connectives are eliminated. Specifically, we find that the
discourse relations expressed by some explicit instances will change when
connectives disappear. Unlike previous work manually analyzing a few examples,
we present empirical evidence at the corpus level to prove the existence of
such shift. Then, we analyze why label shift occurs by considering factors such
as the syntactic role played by connectives, ambiguity of connectives, and
more. Finally, we investigate two strategies to mitigate the label shift:
filtering out noisy data and joint learning with connectives. Experiments on
PDTB 2.0, PDTB 3.0, and the GUM dataset demonstrate that classifiers trained
with our strategies outperform strong baselines. | cs.CL | What Causes the Failure of Explicit to Implicit Discourse Relation Recognition? |
2024-04-01T09:02:12Z | http://arxiv.org/pdf/2404.00998v1 | http://arxiv.org/abs/2404.00998v1 | Zilong Wang, Xufang Luo, Xinyang Jiang, Dongsheng Li, Lili Qiu | 2024-04-01T09:02:12Z | cs.CL, cs.AI | Evaluating generated radiology reports is crucial for the development of
radiology AI, but existing metrics fail to reflect the task's clinical
requirements. This study proposes a novel evaluation framework using large
language models (LLMs) to compare radiology reports for assessment. We compare
the performance of various LLMs and demonstrate that, when using GPT-4, our
proposed metric achieves evaluation consistency close to that of radiologists.
Furthermore, to reduce costs and improve accessibility, making this method
practical, we construct a dataset using LLM evaluation results and perform
knowledge distillation to train a smaller model. The distilled model achieves
evaluation capabilities comparable to GPT-4. Our framework and distilled model
offer an accessible and efficient evaluation method for radiology report
generation, facilitating the development of more clinically relevant models.
The model will be further open-sourced and accessible. | cs.CL | LLM-RadJudge: Achieving Radiologist-Level Evaluation for X-Ray Report Generation |
2024-04-01T08:35:56Z | http://arxiv.org/pdf/2404.00990v1 | http://arxiv.org/abs/2404.00990v1 | Weicong Qin, Zhongxiang Sun | 2024-04-01T08:35:56Z | cs.CL | With the advancement of Artificial Intelligence (AI) and Large Language
Models (LLMs), there is a profound transformation occurring in the realm of
natural language processing tasks within the legal domain. The capabilities of
LLMs are increasingly demonstrating unique roles in the legal sector, bringing
both distinctive benefits and various challenges. This survey delves into the
synergy between LLMs and the legal system, such as their applications in tasks
like legal text comprehension, case retrieval, and analysis. Furthermore, this
survey highlights key challenges faced by LLMs in the legal domain, including
bias, interpretability, and ethical considerations, as well as how researchers
are addressing these issues. The survey showcases the latest advancements in
fine-tuned legal LLMs tailored for various legal systems, along with legal
datasets available for fine-tuning LLMs in various languages. Additionally, it
proposes directions for future research and development. | cs.CL | Exploring the Nexus of Large Language Models and Legal Systems: A Short Survey |
2024-04-01T07:49:11Z | http://arxiv.org/pdf/2404.00978v1 | http://arxiv.org/abs/2404.00978v1 | Hang Zhou, Chenglong Wang, Yimin Hu, Tong Xiao, Chunliang Zhang, Jingbo Zhu | 2024-04-01T07:49:11Z | cs.CL | Reinforcement learning with human feedback for aligning large language models
(LLMs) trains a reward model typically using ranking loss with comparison
pairs.However, the training procedure suffers from an inherent problem: the
uncontrolled scaling of reward scores during reinforcement learning due to the
lack of constraints while training the reward model.This paper proposes a Prior
Constraints-based Reward Model (namely PCRM) training method to mitigate this
problem. PCRM incorporates prior constraints, specifically, length ratio and
cosine similarity between outputs of each comparison pair, during reward model
training to regulate optimization magnitude and control score margins. We
comprehensively evaluate PCRM by examining its rank correlation with human
preferences and its effectiveness in aligning LLMs via RL. Experimental results
demonstrate that PCRM significantly improves alignment performance by
effectively constraining reward score scaling. As another bonus, our method is
easily integrated into arbitrary rank-based alignment methods, such as direct
preference optimization, and can yield consistent improvement. | cs.CL | Prior Constraints-based Reward Model Training for Aligning Large Language Models |
2024-04-01T07:35:15Z | http://arxiv.org/pdf/2404.01353v1 | http://arxiv.org/abs/2404.01353v1 | Achintya Kundu, Fabian Lim, Aaron Chew, Laura Wynter, Penny Chong, Rhui Dih Lee | 2024-04-01T07:35:15Z | cs.LG, cs.AI, cs.CL | Supernet training of LLMs is of great interest in industrial applications as
it confers the ability to produce a palette of smaller models at constant cost,
regardless of the number of models (of different size / latency) produced. We
propose a new method called Multistage Low-rank Fine-tuning of
Super-transformers (MLFS) for parameter-efficient supernet training. We show
that it is possible to obtain high-quality encoder models that are suitable for
commercial edge applications, and that while decoder-only models are resistant
to a comparable degree of compression, decoders can be effectively sliced for a
significant reduction in training time. | cs.LG | Efficiently Distilling LLMs for Edge Applications |
2024-04-01T06:50:52Z | http://arxiv.org/pdf/2404.03683v1 | http://arxiv.org/abs/2404.03683v1 | Kanishk Gandhi, Denise Lee, Gabriel Grand, Muxin Liu, Winson Cheng, Archit Sharma, Noah D. Goodman | 2024-04-01T06:50:52Z | cs.LG, cs.AI, cs.CL | Language models are rarely shown fruitful mistakes while training. They then
struggle to look beyond the next token, suffering from a snowballing of errors
and struggling to predict the consequence of their actions several steps ahead.
In this paper, we show how language models can be taught to search by
representing the process of search in language, as a flattened string -- a
stream of search (SoS). We propose a unified language for search that captures
an array of different symbolic search strategies. We demonstrate our approach
using the simple yet difficult game of Countdown, where the goal is to combine
input numbers with arithmetic operations to reach a target number. We pretrain
a transformer-based language model from scratch on a dataset of streams of
search generated by heuristic solvers. We find that SoS pretraining increases
search accuracy by 25% over models trained to predict only the optimal search
trajectory. We further finetune this model with two policy improvement methods:
Advantage-Induced Policy Alignment (APA) and Self-Taught Reasoner (STaR). The
finetuned SoS models solve 36% of previously unsolved problems, including
problems that cannot be solved by any of the heuristic solvers. Our results
indicate that language models can learn to solve problems via search,
self-improve to flexibly use different search strategies, and potentially
discover new ones. | cs.LG | Stream of Search (SoS): Learning to Search in Language |
2024-04-01T06:25:47Z | http://arxiv.org/pdf/2404.00950v1 | http://arxiv.org/abs/2404.00950v1 | Renhua Gu, Xiangfeng Meng | 2024-04-01T06:25:47Z | cs.CL | SemEval-2024 Task 8 provides a challenge to detect human-written and
machine-generated text. There are 3 subtasks for different detection scenarios.
This paper proposes a system that mainly deals with Subtask B. It aims to
detect if given full text is written by human or is generated by a specific
Large Language Model (LLM), which is actually a multi-class text classification
task. Our team AISPACE conducted a systematic study of fine-tuning
transformer-based models, including encoderonly, decoder-only and
encoder-decoder models. We compared their performance on this task and
identified that encoder-only models performed exceptionally well. We also
applied a weighted Cross Entropy loss function to address the issue of data
imbalance of different class samples. Additionally, we employed softvoting
strategy over multi-models ensemble to enhance the reliability of our
predictions. Our system ranked top 1 in Subtask B, which sets a
state-of-the-art benchmark for this new challenge. | cs.CL | AISPACE at SemEval-2024 task 8: A Class-balanced Soft-voting System for Detecting Multi-generator Machine-generated Text |
2024-04-01T06:03:39Z | http://arxiv.org/pdf/2404.00943v1 | http://arxiv.org/abs/2404.00943v1 | Jihoo Kim, Wonho Song, Dahyun Kim, Yunsu Kim, Yungi Kim, Chanjun Park | 2024-04-01T06:03:39Z | cs.CL, cs.AI | This paper introduces Evalverse, a novel library that streamlines the
evaluation of Large Language Models (LLMs) by unifying disparate evaluation
tools into a single, user-friendly framework. Evalverse enables individuals
with limited knowledge of artificial intelligence to easily request LLM
evaluations and receive detailed reports, facilitated by an integration with
communication platforms like Slack. Thus, Evalverse serves as a powerful tool
for the comprehensive assessment of LLMs, offering both researchers and
practitioners a centralized and easily accessible evaluation framework.
Finally, we also provide a demo video for Evalverse, showcasing its
capabilities and implementation in a two-minute format. | cs.CL | Evalverse: Unified and Accessible Library for Large Language Model Evaluation |
2024-04-01T06:01:17Z | http://arxiv.org/pdf/2404.00942v1 | http://arxiv.org/abs/2404.00942v1 | Xiaoze Liu, Feijie Wu, Tianyang Xu, Zhuo Chen, Yichi Zhang, Xiaoqian Wang, Jing Gao | 2024-04-01T06:01:17Z | cs.CL, cs.AI, cs.LG | The advent of Large Language Models (LLMs) has significantly transformed the
AI landscape, enhancing machine learning and AI capabilities. Factuality issue
is a critical concern for LLMs, as they may generate factually incorrect
responses. In this paper, we propose GraphEval to evaluate an LLM's performance
using a substantially large test dataset. Specifically, the test dataset is
retrieved from a large knowledge graph with more than 10 million facts without
expensive human efforts. Unlike conventional methods that evaluate LLMs based
on generated responses, GraphEval streamlines the evaluation process by
creating a judge model to estimate the correctness of the answers given by the
LLM. Our experiments demonstrate that the judge model's factuality assessment
aligns closely with the correctness of the LLM's generated outputs, while also
substantially reducing evaluation costs. Besides, our findings offer valuable
insights into LLM performance across different metrics and highlight the
potential for future improvements in ensuring the factual integrity of LLM
outputs. The code is publicly available at https://github.com/xz-liu/GraphEval. | cs.CL | Evaluating the Factuality of Large Language Models using Large-Scale Knowledge Graphs |
2024-04-01T05:50:56Z | http://arxiv.org/pdf/2404.00938v2 | http://arxiv.org/abs/2404.00938v2 | Zhonghao Shi, Ellen Landrum, Amy O' Connell, Mina Kian, Leticia Pinto-Alva, Kaleen Shrestha, Xiaoyuan Zhu, Maja J Matarić | 2024-04-05T21:55:47Z | cs.HC, cs.CL, cs.CV, cs.RO | Socially assistive robots (SARs) have shown great success in providing
personalized cognitive-affective support for user populations with special
needs such as older adults, children with autism spectrum disorder (ASD), and
individuals with mental health challenges. The large body of work on SAR
demonstrates its potential to provide at-home support that complements
clinic-based interventions delivered by mental health professionals, making
these interventions more effective and accessible. However, there are still
several major technical challenges that hinder SAR-mediated interactions and
interventions from reaching human-level social intelligence and efficacy. With
the recent advances in large language models (LLMs), there is an increased
potential for novel applications within the field of SAR that can significantly
expand the current capabilities of SARs. However, incorporating LLMs introduces
new risks and ethical concerns that have not yet been encountered, and must be
carefully be addressed to safely deploy these more advanced systems. In this
work, we aim to conduct a brief survey on the use of LLMs in SAR technologies,
and discuss the potentials and risks of applying LLMs to the following three
major technical challenges of SAR: 1) natural language dialog; 2) multimodal
understanding; 3) LLMs as robot policies. | cs.HC | How Can Large Language Models Enable Better Socially Assistive Human-Robot Interaction: A Brief Survey |
2024-04-01T05:39:36Z | http://arxiv.org/pdf/2404.00934v2 | http://arxiv.org/abs/2404.00934v2 | Zhenyu Hou, Yilin Niu, Zhengxiao Du, Xiaohan Zhang, Xiao Liu, Aohan Zeng, Qinkai Zheng, Minlie Huang, Hongning Wang, Jie Tang, Yuxiao Dong | 2024-04-03T17:04:06Z | cs.CL | ChatGLM is a free-to-use AI service powered by the ChatGLM family of large
language models (LLMs). In this paper, we present the ChatGLM-RLHF pipeline --
a reinforcement learning from human feedback (RLHF) system -- designed to
enhance ChatGLM's alignment with human preferences. ChatGLM-RLHF encompasses
three major components: the collection of human preference data, the training
of the reward model, and the optimization of policies. Throughout the process
of integrating ChatGLM-RLHF into production, we encountered and addressed
several unprecedented challenges. We introduce the strategies to mitigate
reward variance for stabilized large-scale training, implement model
parallelism with fused gradient-descent, and design regularization constraints
to avoid catastrophic forgetting in LLMs. Experiments show that ChatGLM-RLHF
brings significant improvements in alignment tasks compared to the supervised
fine-tuned (SFT) version of ChatGLM. For instance, it achieves on average 15\%
more wins against ChatGLM-SFT in Chinese alignment tasks. The work presents our
practices of aligning LLMs with human preferences, offering insights into the
challenges and solutions in RLHF implementations. | cs.CL | ChatGLM-RLHF: Practices of Aligning Large Language Models with Human Feedback |
2024-04-01T05:19:34Z | http://arxiv.org/pdf/2404.00930v1 | http://arxiv.org/abs/2404.00930v1 | Ji-Eun Han, Jun-Seok Koh, Hyeon-Tae Seo, Du-Seong Chang, Kyung-Ah Sohn | 2024-04-01T05:19:34Z | cs.CL | We present a novel end-to-end personality-based synthetic dialogue data
generation pipeline, specifically designed to elicit responses from large
language models via prompting. We design the prompts to generate more
human-like dialogues considering real-world scenarios when users engage with
chatbots. We introduce PSYDIAL, the first Korean dialogue dataset focused on
personality-based dialogues, curated using our proposed pipeline. Notably, we
focus on the Extraversion dimension of the Big Five personality model in our
research. Experimental results indicate that while pre-trained models and those
fine-tuned with a chit-chat dataset struggle to generate responses reflecting
personality, models trained with PSYDIAL show significant improvements. The
versatility of our pipeline extends beyond dialogue tasks, offering potential
for other non-dialogue related applications. This research opens doors for more
nuanced, personality-driven conversational AI in Korean and potentially other
languages. Our code is publicly available at
https://github.com/jiSilverH/psydial. | cs.CL | PSYDIAL: Personality-based Synthetic Dialogue Generation using Large Language Models |
2024-04-01T05:13:56Z | http://arxiv.org/pdf/2404.00929v2 | http://arxiv.org/abs/2404.00929v2 | Yuemei Xu, Ling Hu, Jiayi Zhao, Zihan Qiu, Yuqi Ye, Hanwen Gu | 2024-06-06T16:04:15Z | cs.CL, cs.AI | Based on the foundation of Large Language Models (LLMs), Multilingual Large
Language Models (MLLMs) have been developed to address the challenges of
multilingual natural language processing tasks, hoping to achieve knowledge
transfer from high-resource to low-resource languages. However, significant
limitations and challenges still exist, such as language imbalance,
multilingual alignment, and inherent bias. In this paper, we aim to provide a
comprehensive analysis of MLLMs, delving deeply into discussions surrounding
these critical issues. First of all, we start by presenting an overview of
MLLMs, covering their evolution, key techniques, and multilingual capacities.
Secondly, we explore widely utilized multilingual corpora for MLLMs' training
and multilingual datasets oriented for downstream tasks that are crucial for
enhancing the cross-lingual capability of MLLMs. Thirdly, we survey the
existing studies on multilingual representations and investigate whether the
current MLLMs can learn a universal language representation. Fourthly, we
discuss bias on MLLMs including its category and evaluation metrics, and
summarize the existing debiasing techniques. Finally, we discuss existing
challenges and point out promising research directions. By demonstrating these
aspects, this paper aims to facilitate a deeper understanding of MLLMs and
their potentiality in various domains. | cs.CL | A Survey on Multilingual Large Language Models: Corpora, Alignment, and Bias |
2024-04-01T05:07:13Z | http://arxiv.org/pdf/2404.00925v1 | http://arxiv.org/abs/2404.00925v1 | Jia Gong, Lin Geng Foo, Yixuan He, Hossein Rahmani, Jun Liu | 2024-04-01T05:07:13Z | cs.CV, cs.CL | Sign Language Translation (SLT) is a challenging task that aims to translate
sign videos into spoken language. Inspired by the strong translation
capabilities of large language models (LLMs) that are trained on extensive
multilingual text corpora, we aim to harness off-the-shelf LLMs to handle SLT.
In this paper, we regularize the sign videos to embody linguistic
characteristics of spoken language, and propose a novel SignLLM framework to
transform sign videos into a language-like representation for improved
readability by off-the-shelf LLMs. SignLLM comprises two key modules: (1) The
Vector-Quantized Visual Sign module converts sign videos into a sequence of
discrete character-level sign tokens, and (2) the Codebook Reconstruction and
Alignment module converts these character-level tokens into word-level sign
representations using an optimal transport formulation. A sign-text alignment
loss further bridges the gap between sign and text tokens, enhancing semantic
compatibility. We achieve state-of-the-art gloss-free results on two
widely-used SLT benchmarks. | cs.CV | LLMs are Good Sign Language Translators |
2024-04-01T04:39:44Z | http://arxiv.org/pdf/2404.00914v1 | http://arxiv.org/abs/2404.00914v1 | Yuanhao Zeng, Min Wang, Yihang Wang, Yingxia Shao | 2024-04-01T04:39:44Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) have excelled in various tasks but perform
better in high-resource scenarios, which presents challenges in low-resource
scenarios. Data scarcity and the inherent difficulty of adapting LLMs to
specific tasks compound the challenge. To address the twin hurdles, we
introduce \textbf{Leverage Learning}. We present a streamlined implement of
this methodology called Token-Efficient Leverage Learning (TELL). TELL
showcases the potential of Leverage Learning, demonstrating effectiveness
across various LLMs and low-resource tasks, ranging from $10^4$ to $10^6$
tokens. It reduces task data requirements by up to nearly an order of magnitude
compared to conventional Supervised Fine-Tuning (SFT) while delivering
competitive performance. With the same amount of task data, TELL leads in
improving task performance compared to SFT. We discuss the mechanism of
Leverage Learning, suggesting it aligns with quantization hypothesis and
explore its promising potential through empirical testing. | cs.CL | Token-Efficient Leverage Learning in Large Language Models |
2024-04-01T04:39:21Z | http://arxiv.org/pdf/2404.00913v1 | http://arxiv.org/abs/2404.00913v1 | Bo Zou, Chao Yang, Yu Qiao, Chengbin Quan, Youjian Zhao | 2024-04-01T04:39:21Z | cs.CV, cs.AI, cs.CL | Existing methods to fine-tune LLMs, like Adapter, Prefix-tuning, and LoRA,
which introduce extra modules or additional input sequences to inject new
skills or knowledge, may compromise the innate abilities of LLMs. In this
paper, we propose LLaMA-Excitor, a lightweight method that stimulates the LLMs'
potential to better follow instructions by gradually paying more attention to
worthwhile information. Specifically, the LLaMA-Excitor does not directly
change the intermediate hidden state during the self-attention calculation of
the transformer structure. We designed the Excitor block as a bypass module for
the similarity score computation in LLMs' self-attention to reconstruct keys
and change the importance of values by learnable prompts. LLaMA-Excitor ensures
a self-adaptive allocation of additional attention to input instructions, thus
effectively preserving LLMs' pre-trained knowledge when fine-tuning LLMs on
low-quality instruction-following datasets. Furthermore, we unify the modeling
of multi-modal tuning and language-only tuning, extending LLaMA-Excitor to a
powerful visual instruction follower without the need for complex multi-modal
alignment. Our proposed approach is evaluated in language-only and multi-modal
tuning experimental scenarios. Notably, LLaMA-Excitor is the only method that
maintains basic capabilities while achieving a significant improvement (+6%) on
the MMLU benchmark. In the visual instruction tuning, we achieve a new
state-of-the-art image captioning performance of 157.5 CIDEr on MSCOCO, and a
comparable performance (88.39%) on ScienceQA to cutting-edge models with more
parameters and extensive vision-language pertaining. | cs.CV | LLaMA-Excitor: General Instruction Tuning via Indirect Feature Interaction |
2024-04-01T03:59:12Z | http://arxiv.org/pdf/2406.15362v1 | http://arxiv.org/abs/2406.15362v1 | Nuredin Ali, Charles Chuankai Zhang, Ned Mayo, Stevie Chancellor | 2024-04-01T03:59:12Z | cs.CL | Social media data has been used for detecting users with mental disorders,
such as depression. Despite the global significance of cross-cultural
representation and its potential impact on model performance, publicly
available datasets often lack crucial metadata related to this aspect. In this
work, we evaluate the generalization of benchmark datasets to build AI models
on cross-cultural Twitter data. We gather a custom geo-located Twitter dataset
of depressed users from seven countries as a test dataset. Our results show
that depression detection models do not generalize globally. The models perform
worse on Global South users compared to Global North. Pre-trained language
models achieve the best generalization compared to Logistic Regression, though
still show significant gaps in performance on depressed and non-Western users.
We quantify our findings and provide several actionable suggestions to mitigate
this issue. | cs.CL | Diverse Perspectives, Divergent Models: Cross-Cultural Evaluation of Depression Detection on Twitter |
2024-04-01T03:54:42Z | http://arxiv.org/pdf/2404.00899v1 | http://arxiv.org/abs/2404.00899v1 | Xiaoyan Qu, Xiangfeng Meng | 2024-04-01T03:54:42Z | cs.CL | With the increasing prevalence of text generated by large language models
(LLMs), there is a growing concern about distinguishing between LLM-generated
and human-written texts in order to prevent the misuse of LLMs, such as the
dissemination of misleading information and academic dishonesty. Previous
research has primarily focused on classifying text as either entirely
human-written or LLM-generated, neglecting the detection of mixed texts that
contain both types of content. This paper explores LLMs' ability to identify
boundaries in human-written and machine-generated mixed texts. We approach this
task by transforming it into a token classification problem and regard the
label turning point as the boundary. Notably, our ensemble model of LLMs
achieved first place in the 'Human-Machine Mixed Text Detection' sub-task of
the SemEval'24 Competition Task 8. Additionally, we investigate factors that
influence the capability of LLMs in detecting boundaries within mixed texts,
including the incorporation of extra layers on top of LLMs, combination of
segmentation loss, and the impact of pretraining. Our findings aim to provide
valuable insights for future research in this area. | cs.CL | TM-TREK at SemEval-2024 Task 8: Towards LLM-Based Automatic Boundary Detection for Human-Machine Mixed Text |
2024-04-01T03:25:06Z | http://arxiv.org/pdf/2404.00884v1 | http://arxiv.org/abs/2404.00884v1 | Wei He, Shichun Liu, Jun Zhao, Yiwen Ding, Yi Lu, Zhiheng Xi, Tao Gui, Qi Zhang, Xuanjing Huang | 2024-04-01T03:25:06Z | cs.CL, cs.AI | Large language models (LLMs) have shown promising abilities of in-context
learning (ICL), adapting swiftly to new tasks with only few-shot
demonstrations. However, current few-shot methods heavily depend on
high-quality, query-specific demos, which are often lacking. When faced with
out-of-demonstration (OOD) queries, methods that rely on hand-crafted demos or
external retrievers might fail. To bridge the gap between limited demos and OOD
queries, we propose Self-Demos, a novel prompting method that elicits the
inherent generalizability in LLMs by query-aware demo generation. The generated
demos strategically interpolate between existing demos and the given query,
transforming the query from OOD to ID. To evaluate the effectiveness of our
approach, we manually constructed OOD-Toolset, a dataset in the tool-using
scenario with over 300 real-world APIs and 1000 instances, each consisting of
three tool-use cases as demos and an OOD query. Thorough experiments on our
dataset and two public math benchmarks have shown that our method can
outperform state-of-the-art baselines in the OOD setting. Moreover, we conduct
a range of analyses to validate Self-Demos's generalization and provide more
insights. | cs.CL | Self-Demos: Eliciting Out-of-Demonstration Generalizability in Large Language Models |
2024-04-01T02:04:44Z | http://arxiv.org/pdf/2404.00862v1 | http://arxiv.org/abs/2404.00862v1 | Lung-Chuan Chen, Zong-Ru Li | 2024-04-01T02:04:44Z | cs.CL, cs.AI | Large language models (LLMs) have demonstrated exceptional performance in
various NLP applications. However, the majority of existing open-source LLMs
are pre-trained primarily on English data and little part of other languages.
This deficiency in multilingual training data results in suboptimal performance
when applied to languages with fewer available resources. Furthermore,
enhancing the performance of LLMs on low-resource languages by full-parameter
fine-tuning with additional data requires substantial computational resources,
posing computational barriers for research organizations and individual
researchers. Consequently, several techniques such as parameter-efficient
tuning and advanced embedding initialization have been proposed to address
these challenges. In this work, we combine them to facilitate cross-lingual
transfer on English-dominated open-source LLM. To effectively enhance the
model's proficiency in Traditional Chinese, we conduct secondary pre-training
on Llama 2 7B with Traditional Chinese data by leveraging QLoRA and our
proposed zip-tie embedding initialization. The resulting model called Bailong,
which stands for Bilingual trAnsfer learnIng based on qLOra and zip-tie
embeddiNG. We present Bailong-instruct 7B, a fine-tuned version of Bailong 7B
optimized for multi-turn dialogue scenarios. Recognizing the inadequacy of
benchmark datasets in Traditional Chinese, we further introduce Bailong-bench
to assess the alignment of models with human preferences and the capability to
follow instructions in both Traditional Chinese and English tasks. In our
evaluation, Bailong-instruct 7B exhibits competitive performance on
Bailong-bench and other benchmark datasets when compared to other open-source
models of similar or even larger parameter sizes. Bailong-instruct 7B and
Bailong-bench are publicly available with the aim of empowering the community
to build upon our efforts. | cs.CL | Bailong: Bilingual Transfer Learning based on QLoRA and Zip-tie Embedding |
2024-04-01T02:01:28Z | http://arxiv.org/pdf/2404.00859v1 | http://arxiv.org/abs/2404.00859v1 | Wilson Wu, John X. Morris, Lionel Levine | 2024-04-01T02:01:28Z | cs.LG, cs.CL | Do transformers "think ahead" during inference at a given position? It is
known transformers prepare information in the hidden states of the forward pass
at $t$ that is then used in future forward passes $t+\tau$. We posit two
explanations for this phenomenon: pre-caching, in which off-diagonal gradient
terms present in training result in the model computing features at $t$
irrelevant to the present inference task but useful for the future, and
breadcrumbs, in which features most relevant to time step $t$ are already the
same as those that would most benefit inference at time $t+\tau$. We test these
hypotheses by training language models without propagating gradients to past
timesteps, a scheme we formalize as myopic training. In a synthetic data
setting, we find clear evidence for pre-caching. In the autoregressive language
modeling setting, our experiments are more suggestive of the breadcrumbs
hypothesis. | cs.LG | Do language models plan ahead for future tokens? |
2024-03-31T23:48:50Z | http://arxiv.org/pdf/2404.00829v1 | http://arxiv.org/abs/2404.00829v1 | Anneliese Brei, Chao Zhao, Snigdha Chaturvedi | 2024-03-31T23:48:50Z | cs.CL | Human writers often bookend their writing with ending sentences that relate
back to the beginning sentences in order to compose a satisfying narrative that
"closes the loop." Motivated by this observation, we propose RENarGen, a
controllable story-generation paradigm that generates narratives by ensuring
the first and last sentences are related and then infilling the middle
sentences. Our contributions include an initial exploration of how various
methods of bookending from Narratology affect language modeling for stories.
Automatic and human evaluations indicate RENarGen produces better stories with
more narrative closure than current autoregressive models. | cs.CL | Returning to the Start: Generating Narratives with Related Endpoints |
2024-03-31T23:46:51Z | http://arxiv.org/pdf/2404.00828v1 | http://arxiv.org/abs/2404.00828v1 | Zhuotong Chen, Zihu Wang, Yifan Yang, Qianxiao Li, Zheng Zhang | 2024-03-31T23:46:51Z | cs.CL | Despite the effectiveness of deep neural networks in numerous natural
language processing applications, recent findings have exposed the
vulnerability of these language models when minor perturbations are introduced.
While appearing semantically indistinguishable to humans, these perturbations
can significantly reduce the performance of well-trained language models,
raising concerns about the reliability of deploying them in safe-critical
situations. In this work, we construct a computationally efficient self-healing
process to correct undesired model behavior during online inference when
perturbations are applied to input data. This is formulated as a trajectory
optimization problem in which the internal states of the neural network layers
are automatically corrected using a PID (Proportional-Integral-Derivative)
control mechanism. The P controller targets immediate state adjustments, while
the I and D controllers consider past states and future dynamical trends,
respectively. We leverage the geometrical properties of the training data to
design effective linear PID controllers. This approach reduces the
computational cost to that of using just the P controller, instead of the full
PID control. Further, we introduce an analytical method for approximating the
optimal control solutions, enhancing the real-time inference capabilities of
this controlled system. Moreover, we conduct a theoretical error analysis of
the analytic solution in a simplified setting. The proposed PID control-based
self-healing is a low cost framework that improves the robustness of
pre-trained large language models, whether standard or robustly trained,
against a wide range of perturbations. A detailed implementation can be found
in:https://github.com/zhuotongchen/PID-Control-Based-Self-Healing-to-Improve-the-Robustness-of-Large-Language-Models. | cs.CL | PID Control-Based Self-Healing to Improve the Robustness of Large Language Models |
2024-03-31T23:37:18Z | http://arxiv.org/pdf/2404.00826v2 | http://arxiv.org/abs/2404.00826v2 | Yujuan Fu, Giridhar Kaushik Ramachandran, Nicholas J Dobbins, Namu Park, Michael Leu, Abby R. Rosenberg, Kevin Lybarger, Fei Xia, Ozlem Uzuner, Meliha Yetisgen | 2024-04-04T04:21:34Z | cs.CL | Social determinants of health (SDoH) play a critical role in shaping health
outcomes, particularly in pediatric populations where interventions can have
long-term implications. SDoH are frequently studied in the Electronic Health
Record (EHR), which provides a rich repository for diverse patient data. In
this work, we present a novel annotated corpus, the Pediatric Social History
Annotation Corpus (PedSHAC), and evaluate the automatic extraction of detailed
SDoH representations using fine-tuned and in-context learning methods with
Large Language Models (LLMs). PedSHAC comprises annotated social history
sections from 1,260 clinical notes obtained from pediatric patients within the
University of Washington (UW) hospital system. Employing an event-based
annotation scheme, PedSHAC captures ten distinct health determinants to
encompass living and economic stability, prior trauma, education access,
substance use history, and mental health with an overall annotator agreement of
81.9 F1. Our proposed fine-tuning LLM-based extractors achieve high performance
at 78.4 F1 for event arguments. In-context learning approaches with GPT-4
demonstrate promise for reliable SDoH extraction with limited annotated
examples, with extraction performance at 82.3 F1 for event triggers. | cs.CL | Extracting Social Determinants of Health from Pediatric Patient Notes Using Large Language Models: Novel Corpus and Methods |
2024-03-31T22:22:53Z | http://arxiv.org/pdf/2404.01349v1 | http://arxiv.org/abs/2404.01349v1 | Zhibo Chu, Zichong Wang, Wenbin Zhang | 2024-03-31T22:22:53Z | cs.CL, cs.AI | Large Language Models (LLMs) have demonstrated remarkable success across
various domains. However, despite their promising performance in numerous
real-world applications, most of these algorithms lack fairness considerations.
Consequently, they may lead to discriminatory outcomes against certain
communities, particularly marginalized populations, prompting extensive study
in fair LLMs. On the other hand, fairness in LLMs, in contrast to fairness in
traditional machine learning, entails exclusive backgrounds, taxonomies, and
fulfillment techniques. To this end, this survey presents a comprehensive
overview of recent advances in the existing literature concerning fair LLMs.
Specifically, a brief introduction to LLMs is provided, followed by an analysis
of factors contributing to bias in LLMs. Additionally, the concept of fairness
in LLMs is discussed categorically, summarizing metrics for evaluating bias in
LLMs and existing algorithms for promoting fairness. Furthermore, resources for
evaluating bias in LLMs, including toolkits and datasets, are summarized.
Finally, existing research challenges and open questions are discussed. | cs.CL | Fairness in Large Language Models: A Taxonomic Survey |
2024-03-31T20:28:44Z | http://arxiv.org/pdf/2404.00790v1 | http://arxiv.org/abs/2404.00790v1 | Mingyang Wang, Heike Adel, Lukas Lange, Jannik Strötgen, Hinrich Schütze | 2024-03-31T20:28:44Z | cs.LG, cs.CL | Continual learning aims at incrementally acquiring new knowledge while not
forgetting existing knowledge. To overcome catastrophic forgetting, methods are
either rehearsal-based, i.e., store data examples from previous tasks for data
replay, or isolate parameters dedicated to each task. However, rehearsal-based
methods raise privacy and memory issues, and parameter-isolation continual
learning does not consider interaction between tasks, thus hindering knowledge
transfer. In this work, we propose MoCL, a rehearsal-free Modular and
Compositional Continual Learning framework which continually adds new modules
to language models and composes them with existing modules. Experiments on
various benchmarks show that MoCL outperforms state of the art and effectively
facilitates knowledge transfer. | cs.LG | Rehearsal-Free Modular and Compositional Continual Learning for Language Models |
2024-03-31T18:08:37Z | http://arxiv.org/pdf/2404.00758v1 | http://arxiv.org/abs/2404.00758v1 | Josip Jukić, Jan Šnajder | 2024-03-31T18:08:37Z | cs.CL | Enhancing generalization and uncertainty quantification in pre-trained
language models (PLMs) is crucial for their effectiveness and reliability.
Building on machine learning research that established the importance of
robustness for improving generalization, we investigate the role of
representation smoothness, achieved via Jacobian and Hessian regularization, in
enhancing PLM performance. Although such regularization methods have proven
effective in computer vision, their application in natural language processing
(NLP), where PLM inputs are derived from a discrete domain, poses unique
challenges. We introduce a novel two-phase regularization approach, JacHess,
which minimizes the norms of the Jacobian and Hessian matrices within PLM
intermediate representations relative to their inputs. Our evaluation using the
GLUE benchmark demonstrates that JacHess significantly improves in-domain
generalization and calibration in PLMs, outperforming unregularized fine-tuning
and other similar regularization methods. | cs.CL | From Robustness to Improved Generalization and Calibration in Pre-trained Language Models |
2024-03-31T17:47:22Z | http://arxiv.org/pdf/2404.00752v1 | http://arxiv.org/abs/2404.00752v1 | Atsumoto Ohashi, Ukyo Honda, Tetsuro Morimura, Yuu Jinnai | 2024-03-31T17:47:22Z | cs.CL, cs.AI | Minimum Bayes-risk (MBR) decoding has recently gained renewed attention in
text generation. MBR decoding considers texts sampled from a model as
pseudo-references and selects the text with the highest similarity to the
others. Therefore, sampling is one of the key elements of MBR decoding, and
previous studies reported that the performance varies by sampling methods. From
a theoretical standpoint, this performance variation is likely tied to how
closely the samples approximate the true distribution of references. However,
this approximation has not been the subject of in-depth study. In this study,
we propose using anomaly detection to measure the degree of approximation. We
first closely examine the performance variation and then show that previous
hypotheses about samples do not correlate well with the variation, but our
introduced anomaly scores do. The results are the first to empirically support
the link between the performance and the core assumption of MBR decoding. | cs.CL | On the True Distribution Approximation of Minimum Bayes-Risk Decoding |
2024-03-31T17:38:33Z | http://arxiv.org/pdf/2404.00750v1 | http://arxiv.org/abs/2404.00750v1 | Paula Rescala, Manoel Horta Ribeiro, Tiancheng Hu, Robert West | 2024-03-31T17:38:33Z | cs.CL, cs.CY | The remarkable and ever-increasing capabilities of Large Language Models
(LLMs) have raised concerns about their potential misuse for creating
personalized, convincing misinformation and propaganda. To gain insights into
LLMs' persuasive capabilities without directly engaging in experimentation with
humans, we propose studying their performance on the related task of detecting
convincing arguments. We extend a dataset by Durmus & Cardie (2018) with
debates, votes, and user traits and propose tasks measuring LLMs' ability to
(1) distinguish between strong and weak arguments, (2) predict stances based on
beliefs and demographic characteristics, and (3) determine the appeal of an
argument to an individual based on their traits. We show that LLMs perform on
par with humans in these tasks and that combining predictions from different
LLMs yields significant performance gains, even surpassing human performance.
The data and code released with this paper contribute to the crucial ongoing
effort of continuously evaluating and monitoring the rapidly evolving
capabilities and potential impact of LLMs. | cs.CL | Can Language Models Recognize Convincing Arguments? |
2024-03-31T17:33:43Z | http://arxiv.org/pdf/2404.00748v1 | http://arxiv.org/abs/2404.00748v1 | Venelin Kovatchev, Matthew Lease | 2024-03-31T17:33:43Z | cs.CL, cs.AI | In this paper we present an exploratory research on quantifying the impact
that data distribution has on the performance and evaluation of NLP models. We
propose an automated framework that measures the data point distribution across
6 different dimensions: ambiguity, difficulty, discriminability, length, noise,
and perplexity.
We use disproportional stratified sampling to measure how much the data
distribution affects absolute (Acc/F1) and relative (Rank) model performance.
We experiment on 2 different datasets (SQUAD and MNLI) and test a total of 135
different models (125 on SQUAD and 10 on MNLI). We demonstrate that without
explicit control of the data distribution, standard evaluation frameworks are
inconsistent and unreliable. We find that the impact of the data is
statistically significant and is often larger than the impact of changing the
metric.
In a second set of experiments, we demonstrate that the impact of data on
evaluation is not just observable, but also predictable. We propose to use
benchmark transparency as a method for comparing datasets and quantifying the
similarity between them. We find that the ``dataset similarity vector'' can be
used to predict how well a model generalizes out of distribution. | cs.CL | Benchmark Transparency: Measuring the Impact of Data on Evaluation |
2024-03-31T16:54:29Z | http://arxiv.org/pdf/2404.00739v1 | http://arxiv.org/abs/2404.00739v1 | Giuseppe G. A. Celano | 2024-03-31T16:54:29Z | cs.CL | In this article, the beta version 0.1.0 of Opera Graeca Adnotata (OGA), the
largest open-access multilayer corpus for Ancient Greek (AG) is presented. OGA
consists of 1,687 literary works and 34M+ tokens coming from the PerseusDL and
OpenGreekAndLatin GitHub repositories, which host AG texts ranging from about
800 BCE to about 250 CE. The texts have been enriched with seven annotation
layers: (i) tokenization layer; (ii) sentence segmentation layer; (iii)
lemmatization layer; (iv) morphological layer; (v) dependency layer; (vi)
dependency function layer; (vii) Canonical Text Services (CTS) citation layer.
The creation of each layer is described by highlighting the main technical and
annotation-related issues encountered. Tokenization, sentence segmentation, and
CTS citation are performed by rule-based algorithms, while morphosyntactic
annotation is the output of the COMBO parser trained on the data of the Ancient
Greek Dependency Treebank. For the sake of scalability and reusability, the
corpus is released in the standoff formats PAULA XML and its offspring LAULA
XML. | cs.CL | Opera Graeca Adnotata: Building a 34M+ Token Multilayer Corpus for Ancient Greek |
2024-03-31T16:00:41Z | http://arxiv.org/pdf/2404.00727v2 | http://arxiv.org/abs/2404.00727v2 | Ian Porada, Xiyuan Zou, Jackie Chi Kit Cheung | 2024-04-23T00:45:22Z | cs.CL | All state-of-the-art coreference resolution (CR) models involve finetuning a
pretrained language model. Whether the superior performance of one CR model
over another is due to the choice of language model or other factors, such as
the task-specific architecture, is difficult or impossible to determine due to
lack of a standardized experimental setup. To resolve this ambiguity, we
systematically evaluate five CR models and control for certain design decisions
including the pretrained language model used by each. When controlling for
language model size, encoder-based CR models outperform more recent
decoder-based models in terms of both accuracy and inference speed.
Surprisingly, among encoder-based CR models, more recent models are not always
more accurate, and the oldest CR model that we test generalizes the best to
out-of-domain textual genres. We conclude that controlling for the choice of
language model reduces most, but not all, of the increase in F1 score reported
in the past five years. | cs.CL | A Controlled Reevaluation of Coreference Resolution Models |
2024-03-31T15:55:49Z | http://arxiv.org/pdf/2404.00725v1 | http://arxiv.org/abs/2404.00725v1 | Michael Hassid, Tal Remez, Jonas Gehring, Roy Schwartz, Yossi Adi | 2024-03-31T15:55:49Z | cs.SE, cs.AI, cs.CL, cs.LG | It is a common belief that large language models (LLMs) are better than
smaller-sized ones. However, larger models also require significantly more time
and compute during inference. This begs the question: what happens when both
models operate under the same budget? (e.g., compute, run-time). To address
this question, we analyze code generation LLMs of various sizes and make
comparisons such as running a 70B model once vs. generating five outputs from a
13B model and selecting one. Our findings reveal that, in a standard unit-test
setup, the repeated use of smaller models can yield consistent improvements,
with gains of up to 15% across five tasks. On the other hand, in scenarios
where unit-tests are unavailable, a ranking-based selection of candidates from
the smaller model falls short of the performance of a single output from larger
ones. Our results highlight the potential of using smaller models instead of
larger ones, and the importance of studying approaches for ranking LLM outputs. | cs.SE | The Larger the Better? Improved LLM Code-Generation via Budget Reallocation |
2024-03-31T15:13:15Z | http://arxiv.org/pdf/2404.08666v1 | http://arxiv.org/abs/2404.08666v1 | Jesse Atuhurra, Hidetaka Kamigaito | 2024-03-31T15:13:15Z | cs.CL, cs.LG | Natural language processing (NLP) has grown significantly since the advent of
the Transformer architecture. Transformers have given birth to pre-trained
large language models (PLMs). There has been tremendous improvement in the
performance of NLP systems across several tasks. NLP systems are on par or, in
some cases, better than humans at accomplishing specific tasks. However, it
remains the norm that \emph{better quality datasets at the time of pretraining
enable PLMs to achieve better performance, regardless of the task.} The need to
have quality datasets has prompted NLP researchers to continue creating new
datasets to satisfy particular needs. For example, the two top NLP conferences,
ACL and EMNLP, accepted ninety-two papers in 2022, introducing new datasets.
This work aims to uncover the trends and insights mined within these datasets.
Moreover, we provide valuable suggestions to researchers interested in curating
datasets in the future. | cs.CL | Revealing Trends in Datasets from the 2022 ACL and EMNLP Conferences |
2024-03-31T14:32:02Z | http://arxiv.org/pdf/2404.00699v1 | http://arxiv.org/abs/2404.00699v1 | Mathieu Ravaut, Bosheng Ding, Fangkai Jiao, Hailin Chen, Xingxuan Li, Ruochen Zhao, Chengwei Qin, Caiming Xiong, Shafiq Joty | 2024-03-31T14:32:02Z | cs.CL | With the rise of Large Language Models (LLMs) in recent years, new
opportunities are emerging, but also new challenges, and contamination is
quickly becoming critical. Business applications and fundraising in AI have
reached a scale at which a few percentage points gained on popular
question-answering benchmarks could translate into dozens of millions of
dollars, placing high pressure on model integrity. At the same time, it is
becoming harder and harder to keep track of the data that LLMs have seen; if
not impossible with closed-source models like GPT-4 and Claude-3 not divulging
any information on the training set. As a result, contamination becomes a
critical issue: LLMs' performance may not be reliable anymore, as the high
performance may be at least partly due to their previous exposure to the data.
This limitation jeopardizes the entire progress in the field of NLP, yet, there
remains a lack of methods on how to efficiently address contamination, or a
clear consensus on prevention, mitigation and classification of contamination.
In this paper, we survey all recent work on contamination with LLMs, and help
the community track contamination levels of LLMs by releasing an open-source
Python library named LLMSanitize implementing major contamination detection
algorithms, which link is: https://github.com/ntunlp/LLMSanitize. | cs.CL | How Much are LLMs Contaminated? A Comprehensive Survey and the LLMSanitize Library |
2024-03-31T13:30:12Z | http://arxiv.org/pdf/2404.00685v2 | http://arxiv.org/abs/2404.00685v2 | Santiago Cuervo, Ricard Marxer | 2024-04-16T06:46:18Z | eess.AS, cs.AI, cs.CL, cs.NE | Speech Language Models (SLMs) aim to learn language from raw audio, without
textual resources. Despite significant advances, our current models exhibit
weak syntax and semantic abilities. However, if the scaling properties of
neural language models hold for the speech modality, these abilities will
improve as the amount of compute used for training increases. In this paper, we
use models of this scaling behavior to estimate the scale at which our current
methods will yield a SLM with the English proficiency of text-based Large
Language Models (LLMs). We establish a strong correlation between pre-training
loss and downstream syntactic and semantic performance in SLMs and LLMs, which
results in predictable scaling of linguistic performance. We show that the
linguistic performance of SLMs scales up to three orders of magnitude more
slowly than that of text-based LLMs. Additionally, we study the benefits of
synthetic data designed to boost semantic understanding and the effects of
coarser speech tokenization. | eess.AS | Scaling Properties of Speech Language Models |
2024-03-31T13:19:36Z | http://arxiv.org/pdf/2404.00681v1 | http://arxiv.org/abs/2404.00681v1 | Dawei Zhu, Wenhao Wu, Yifan Song, Fangwei Zhu, Ziqiang Cao, Sujian Li | 2024-03-31T13:19:36Z | cs.CL | Coherence evaluation aims to assess the organization and structure of a
discourse, which remains challenging even in the era of large language models.
Due to the scarcity of annotated data, data augmentation is commonly used for
training coherence evaluation models. However, previous augmentations for this
task primarily rely on heuristic rules, lacking designing criteria as guidance.
In this paper, we take inspiration from linguistic theory of discourse
structure, and propose a data augmentation framework named CoUDA. CoUDA breaks
down discourse coherence into global and local aspects, and designs
augmentation strategies for both aspects, respectively. Especially for local
coherence, we propose a novel generative strategy for constructing augmentation
samples, which involves post-pretraining a generative model and applying two
controlling mechanisms to control the difficulty of generated samples. During
inference, CoUDA also jointly evaluates both global and local aspects to
comprehensively assess the overall coherence of a discourse. Extensive
experiments in coherence evaluation show that, with only 233M parameters, CoUDA
achieves state-of-the-art performance in both pointwise scoring and pairwise
ranking tasks, even surpassing recent GPT-3.5 and GPT-4 based metrics. | cs.CL | CoUDA: Coherence Evaluation via Unified Data Augmentation |
2024-03-31T12:44:24Z | http://arxiv.org/pdf/2404.00672v1 | http://arxiv.org/abs/2404.00672v1 | Wenxuan Huang, Yunhang Shen, Jiao Xie, Baochang Zhang, Gaoqi He, Ke Li, Xing Sun, Shaohui Lin | 2024-03-31T12:44:24Z | cs.LG, cs.AI, cs.CL, cs.CV | The remarkable performance of Vision Transformers (ViTs) typically requires
an extremely large training cost. Existing methods have attempted to accelerate
the training of ViTs, yet typically disregard method universality with accuracy
dropping. Meanwhile, they break the training consistency of the original
transformers, including the consistency of hyper-parameters, architecture, and
strategy, which prevents them from being widely applied to different
Transformer networks. In this paper, we propose a novel token growth scheme
Token Expansion (termed ToE) to achieve consistent training acceleration for
ViTs. We introduce an "initialization-expansion-merging" pipeline to maintain
the integrity of the intermediate feature distribution of original
transformers, preventing the loss of crucial learnable information in the
training process. ToE can not only be seamlessly integrated into the training
and fine-tuning process of transformers (e.g., DeiT and LV-ViT), but also
effective for efficient training frameworks (e.g., EfficientTrain), without
twisting the original training hyper-parameters, architecture, and introducing
additional training strategies. Extensive experiments demonstrate that ToE
achieves about 1.3x faster for the training of ViTs in a lossless manner, or
even with performance gains over the full-token training baselines. Code is
available at https://github.com/Osilly/TokenExpansion . | cs.LG | A General and Efficient Training for Transformer via Token Expansion |
2024-03-31T12:01:34Z | http://arxiv.org/pdf/2404.00657v1 | http://arxiv.org/abs/2404.00657v1 | Sumit Soman, Sujoy Roychowdhury | 2024-03-31T12:01:34Z | cs.LG, cs.AI, cs.CL, I.2.7 | Retrieval augmented generation (RAG) for technical documents creates
challenges as embeddings do not often capture domain information. We review
prior art for important factors affecting RAG and perform experiments to
highlight best practices and potential challenges to build RAG systems for
technical documents. | cs.LG | Observations on Building RAG Systems for Technical Documents |
2024-03-31T12:01:32Z | http://arxiv.org/pdf/2404.00656v1 | http://arxiv.org/abs/2404.00656v1 | Shujie Hu, Long Zhou, Shujie Liu, Sanyuan Chen, Hongkun Hao, Jing Pan, Xunying Liu, Jinyu Li, Sunit Sivasankaran, Linquan Liu, Furu Wei | 2024-03-31T12:01:32Z | cs.CL, cs.AI, cs.SD, eess.AS | The recent advancements in large language models (LLMs) have revolutionized
the field of natural language processing, progressively broadening their scope
to multimodal perception and generation. However, effectively integrating
listening capabilities into LLMs poses significant challenges, particularly
with respect to generalizing across varied contexts and executing complex
auditory tasks. In this work, we introduce WavLLM, a robust and adaptive speech
large language model with dual encoders, and a prompt-aware LoRA weight
adapter, optimized by a two-stage curriculum learning approach. Leveraging dual
encoders, we decouple different types of speech information, utilizing a
Whisper encoder to process the semantic content of speech, and a WavLM encoder
to capture the unique characteristics of the speaker's identity. Within the
curriculum learning framework, WavLLM first builds its foundational
capabilities by optimizing on mixed elementary single tasks, followed by
advanced multi-task training on more complex tasks such as combinations of the
elementary tasks. To enhance the flexibility and adherence to different tasks
and instructions, a prompt-aware LoRA weight adapter is introduced in the
second advanced multi-task training stage. We validate the proposed model on
universal speech benchmarks including tasks such as ASR, ST, SV, ER, and also
apply it to specialized datasets like Gaokao English listening comprehension
set for SQA, and speech Chain-of-Thought (CoT) evaluation set. Experiments
demonstrate that the proposed model achieves state-of-the-art performance
across a range of speech tasks on the same model size, exhibiting robust
generalization capabilities in executing complex tasks using CoT approach.
Furthermore, our model successfully completes Gaokao tasks without specialized
training. The codes, models, audio, and Gaokao evaluation set can be accessed
at \url{aka.ms/wavllm}. | cs.CL | WavLLM: Towards Robust and Adaptive Speech Large Language Model |
2024-03-31T09:52:25Z | http://arxiv.org/pdf/2404.01345v1 | http://arxiv.org/abs/2404.01345v1 | Utsha Roy, Mst. Sazia Tahosin, Md. Mahedi Hassan, Taminul Islam, Fahim Imtiaz, Md Rezwane Sadik, Yassine Maleh, Rejwan Bin Sulaiman, Md. Simul Hasan Talukder | 2024-03-31T09:52:25Z | cs.CL, cs.LG | The rise of fake news has made the need for effective detection methods,
including in languages other than English, increasingly important. The study
aims to address the challenges of Bangla which is considered a less important
language. To this end, a complete dataset containing about 50,000 news items is
proposed. Several deep learning models have been tested on this dataset,
including the bidirectional gated recurrent unit (GRU), the long short-term
memory (LSTM), the 1D convolutional neural network (CNN), and hybrid
architectures. For this research, we assessed the efficacy of the model
utilizing a range of useful measures, including recall, precision, F1 score,
and accuracy. This was done by employing a big application. We carry out
comprehensive trials to show the effectiveness of these models in identifying
bogus news in Bangla, with the Bidirectional GRU model having a stunning
accuracy of 99.16%. Our analysis highlights the importance of dataset balance
and the need for continual improvement efforts to a substantial degree. This
study makes a major contribution to the creation of Bangla fake news detecting
systems with limited resources, thereby setting the stage for future
improvements in the detection process. | cs.CL | Enhancing Bangla Fake News Detection Using Bidirectional Gated Recurrent Units and Deep Learning Techniques |
2024-03-31T09:50:39Z | http://arxiv.org/pdf/2404.00629v1 | http://arxiv.org/abs/2404.00629v1 | Lizhi Lin, Honglin Mu, Zenan Zhai, Minghan Wang, Yuxia Wang, Renxi Wang, Junjie Gao, Yixuan Zhang, Wanxiang Che, Timothy Baldwin, Xudong Han, Haonan Li | 2024-03-31T09:50:39Z | cs.CL | Generative models are rapidly gaining popularity and being integrated into
everyday applications, raising concerns over their safety issues as various
vulnerabilities are exposed. Faced with the problem, the field of red teaming
is experiencing fast-paced growth, which highlights the need for a
comprehensive organization covering the entire pipeline and addressing emerging
topics for the community. Our extensive survey, which examines over 120 papers,
introduces a taxonomy of fine-grained attack strategies grounded in the
inherent capabilities of language models. Additionally, we have developed the
searcher framework that unifies various automatic red teaming approaches.
Moreover, our survey covers novel areas including multimodal attacks and
defenses, risks around multilingual models, overkill of harmless queries, and
safety of downstream applications. We hope this survey can provide a systematic
perspective on the field and unlock new areas of research. | cs.CL | Against The Achilles' Heel: A Survey on Red Teaming for Generative Models |
2024-03-31T09:17:34Z | http://arxiv.org/pdf/2404.00620v1 | http://arxiv.org/abs/2404.00620v1 | Deborah N. Jakobi, Daniel G. Krakowczyk, Lena A. Jäger | 2024-03-31T09:17:34Z | cs.CL | Eye-tracking datasets are often shared in the format used by their creators
for their original analyses, usually resulting in the exclusion of data
considered irrelevant to the primary purpose. In order to increase re-usability
of existing eye-tracking datasets for more diverse and initially not considered
use cases, this work advocates a new approach of sharing eye-tracking data.
Instead of publishing filtered and pre-processed datasets, the eye-tracking
data at all pre-processing stages should be published together with data
quality reports. In order to transparently report data quality and enable
cross-dataset comparisons, we develop data quality reporting standards and
metrics that can be automatically applied to a dataset, and integrate them into
the open-source Python package pymovements
(https://github.com/aeye-lab/pymovements). | cs.CL | Reporting Eye-Tracking Data Quality: Towards a New Standard |
2024-03-31T09:04:01Z | http://arxiv.org/pdf/2404.00614v1 | http://arxiv.org/abs/2404.00614v1 | Nathan Cornille, Marie-Francine Moens, Florian Mai | 2024-03-31T09:04:01Z | cs.CL, cs.AI | By training to predict the next token in an unlabeled corpus, large language
models learn to perform many tasks without any labeled data. However, their
next-token-prediction objective arguably limits their performance in scenarios
that require planning, such as writing a coherent article. In this paper, we
train a module for planning the future writing process via a self-supervised
learning objective. By conditioning on generated latent plans, our model
extends the successful language model formula to more abstract planning in an
unsupervised way. Empirically, we demonstrate that our method improves language
modeling performance in general, particularly with respect to the text
structure. Because our framework uses a planner module that is unsupervised and
external to the language model, new planner modules can be trained at large
scale and easily be shared with the community. | cs.CL | Learning to Plan for Language Modeling from Unlabeled Data |
2024-03-31T08:58:54Z | http://arxiv.org/pdf/2404.00610v1 | http://arxiv.org/abs/2404.00610v1 | Chi-Min Chan, Chunpu Xu, Ruibin Yuan, Hongyin Luo, Wei Xue, Yike Guo, Jie Fu | 2024-03-31T08:58:54Z | cs.CL | Large Language Models (LLMs) exhibit remarkable capabilities but are prone to
generating inaccurate or hallucinatory responses. This limitation stems from
their reliance on vast pretraining datasets, making them susceptible to errors
in unseen scenarios. To tackle these challenges, Retrieval-Augmented Generation
(RAG) addresses this by incorporating external, relevant documents into the
response generation process, thus leveraging non-parametric knowledge alongside
LLMs' in-context learning abilities. However, existing RAG implementations
primarily focus on initial input for context retrieval, overlooking the nuances
of ambiguous or complex queries that necessitate further clarification or
decomposition for accurate responses. To this end, we propose learning to
Refine Query for Retrieval Augmented Generation (RQ-RAG) in this paper,
endeavoring to enhance the model by equipping it with capabilities for explicit
rewriting, decomposition, and disambiguation. Our experimental results indicate
that our method, when applied to a 7B Llama2 model, surpasses the previous
state-of-the-art (SOTA) by an average of 1.9\% across three single-hop QA
datasets, and also demonstrates enhanced performance in handling complex,
multi-hop QA datasets. Our code is available at
https://github.com/chanchimin/RQ-RAG. | cs.CL | RQ-RAG: Learning to Refine Queries for Retrieval Augmented Generation |
2024-03-31T08:30:15Z | http://arxiv.org/pdf/2404.00604v1 | http://arxiv.org/abs/2404.00604v1 | Xiao Liu, Xixuan Song, Yuxiao Dong, Jie Tang | 2024-03-31T08:30:15Z | cs.CL, cs.AI, cs.LG | Reinforcement learning from human feedback (RLHF) has been a central
technique for recent large language model (LLM) alignment. However, its heavy
dependence on costly human or LLM-as-Judge preference feedback could stymie its
wider applications. In this work, we introduce Self-Contrast, a feedback-free
large language model alignment method via exploiting extensive self-generated
negatives. With only supervised fine-tuning (SFT) targets, Self-Contrast
leverages the LLM itself to generate massive diverse candidates, and harnesses
a pre-trained embedding model to filter multiple negatives according to text
similarity. Theoretically, we illustrate that in this setting, merely scaling
negative responses can still effectively approximate situations with more
balanced positive and negative preference annotations. Our experiments with
direct preference optimization (DPO) on three datasets show that, Self-Contrast
could consistently outperform SFT and standard DPO training by large margins.
And as the number of self-generated negatives increases, the performance of
Self-Contrast continues to grow. Code and data are available at
https://github.com/THUDM/Self-Contrast. | cs.CL | Extensive Self-Contrast Enables Feedback-Free Language Model Alignment |
2024-03-31T08:14:25Z | http://arxiv.org/pdf/2404.00600v2 | http://arxiv.org/abs/2404.00600v2 | Nicola Fabiano | 2024-04-02T06:05:29Z | cs.CY, cs.AI, cs.CL | The imposing evolution of artificial intelligence systems and, specifically,
of Large Language Models (LLM) makes it necessary to carry out assessments of
their level of risk and the impact they may have in the area of privacy,
personal data protection and at an ethical level, especially on the weakest and
most vulnerable. This contribution addresses human oversight, ethical
oversight, and privacy impact assessment. | cs.CY | AI Act and Large Language Models (LLMs): When critical issues and privacy impact require human and ethical oversight |
2024-03-31T08:10:50Z | http://arxiv.org/pdf/2404.00599v1 | http://arxiv.org/abs/2404.00599v1 | Jia Li, Ge Li, Xuanming Zhang, Yihong Dong, Zhi Jin | 2024-03-31T08:10:50Z | cs.CL, cs.AI, cs.SE | How to evaluate Large Language Models (LLMs) in code generation is an open
question. Existing benchmarks demonstrate poor alignment with real-world code
repositories and are insufficient to evaluate the coding abilities of LLMs.
This paper proposes a new benchmark - EvoCodeBench to address the preceding
problems, which has three primary advances. (1) EvoCodeBench aligns with
real-world repositories in multiple dimensions, e.g., code distributions and
dependency distributions. (2) EvoCodeBench offers comprehensive annotations
(e.g., requirements, reference code, and reference dependencies), and robust
evaluation metrics (e.g., Pass@k and Recall@k). (3) EvoCodeBench is an evolving
benchmark to avoid data leakage. We build an automatic pipeline to update
EvoCodeBench from the latest repositories. We release the first version -
EvoCodeBench-2403, containing 275 samples from 25 real-world repositories.
Based on EvoCodeBench, we propose repository-level code generation and evaluate
10 popular LLMs (e.g., gpt-4, gpt-3.5, DeepSeek Coder, StarCoder 2, CodeLLaMa,
Gemma, and Qwen 1.5). Our experiments reveal the coding abilities of these LLMs
in real-world repositories. For example, the highest Pass@1 of gpt-4 only is
20.73% in our experiments. We also analyze failed cases and summarize the
shortcomings of existing LLMs in EvoCodeBench. We release EvoCodeBench, all
prompts, and LLMs' completions for further community analysis. | cs.CL | EvoCodeBench: An Evolving Code Generation Benchmark Aligned with Real-World Code Repositories |
2024-03-31T08:10:45Z | http://arxiv.org/pdf/2404.01344v1 | http://arxiv.org/abs/2404.01344v1 | T. Y. S. S Santosh, Hassan Sarwat, Ahmed Abdou, Matthias Grabmair | 2024-03-31T08:10:45Z | cs.CL | Rhetorical Role Labeling (RRL) of legal judgments is essential for various
tasks, such as case summarization, semantic search and argument mining.
However, it presents challenges such as inferring sentence roles from context,
interrelated roles, limited annotated data, and label imbalance. This study
introduces novel techniques to enhance RRL performance by leveraging knowledge
from semantically similar instances (neighbours). We explore inference-based
and training-based approaches, achieving remarkable improvements in challenging
macro-F1 scores. For inference-based methods, we explore interpolation
techniques that bolster label predictions without re-training. While in
training-based methods, we integrate prototypical learning with our novel
discourse-aware contrastive method that work directly on embedding spaces.
Additionally, we assess the cross-domain applicability of our methods,
demonstrating their effectiveness in transferring knowledge across diverse
legal domains. | cs.CL | Mind Your Neighbours: Leveraging Analogous Instances for Rhetorical Role Labeling for Legal Documents |
2024-03-31T08:06:54Z | http://arxiv.org/pdf/2404.00596v1 | http://arxiv.org/abs/2404.00596v1 | T. Y. S. S Santosh, Rashid Gustav Haddad, Matthias Grabmair | 2024-03-31T08:06:54Z | cs.CL, cs.IR | In common law jurisdictions, legal practitioners rely on precedents to
construct arguments, in line with the doctrine of \emph{stare decisis}. As the
number of cases grow over the years, prior case retrieval (PCR) has garnered
significant attention. Besides lacking real-world scale, existing PCR datasets
do not simulate a realistic setting, because their queries use complete case
documents while only masking references to prior cases. The query is thereby
exposed to legal reasoning not yet available when constructing an argument for
an undecided case as well as spurious patterns left behind by citation masks,
potentially short-circuiting a comprehensive understanding of case facts and
legal principles. To address these limitations, we introduce a PCR dataset
based on judgements from the European Court of Human Rights (ECtHR), which
explicitly separate facts from arguments and exhibit precedential practices,
aiding us to develop this PCR dataset to foster systems' comprehensive
understanding. We benchmark different lexical and dense retrieval approaches
with various negative sampling strategies, adapting them to deal with long text
sequences using hierarchical variants. We found that difficulty-based negative
sampling strategies were not effective for the PCR task, highlighting the need
for investigation into domain-specific difficulty criteria. Furthermore, we
observe performance of the dense models degrade with time and calls for further
research into temporal adaptation of retrieval models. Additionally, we assess
the influence of different views , Halsbury's and Goodhart's, in practice in
ECtHR jurisdiction using PCR task. | cs.CL | ECtHR-PCR: A Dataset for Precedent Understanding and Prior Case Retrieval in the European Court of Human Rights |
2024-03-31T08:03:39Z | http://arxiv.org/pdf/2404.00595v1 | http://arxiv.org/abs/2404.00595v1 | T. Y. S. S Santosh, Elvin Quero Hernandez, Matthias Grabmair | 2024-03-31T08:03:39Z | cs.CL, cs.IR | Legal professionals often grapple with navigating lengthy legal judgements to
pinpoint information that directly address their queries. This paper focus on
this task of extracting relevant paragraphs from legal judgements based on the
query. We construct a specialized dataset for this task from the European Court
of Human Rights (ECtHR) using the case law guides. We assess the performance of
current retrieval models in a zero-shot way and also establish fine-tuning
benchmarks using various models. The results highlight the significant gap
between fine-tuned and zero-shot performance, emphasizing the challenge of
handling distribution shift in the legal domain. We notice that the legal
pre-training handles distribution shift on the corpus side but still struggles
on query side distribution shift, with unseen legal queries. We also explore
various Parameter Efficient Fine-Tuning (PEFT) methods to evaluate their
practicality within the context of information retrieval, shedding light on the
effectiveness of different PEFT methods across diverse configurations with
pre-training and model architectures influencing the choice of PEFT method. | cs.CL | Query-driven Relevant Paragraph Extraction from Legal Judgments |
2024-03-31T08:00:40Z | http://arxiv.org/pdf/2404.00594v1 | http://arxiv.org/abs/2404.00594v1 | T. Y. S. S Santosh, Mahmoud Aly, Matthias Grabmair | 2024-03-31T08:00:40Z | cs.CL | Legal professionals frequently encounter long legal judgments that hold
critical insights for their work. While recent advances have led to automated
summarization solutions for legal documents, they typically provide generic
summaries, which may not meet the diverse information needs of users. To
address this gap, we introduce LexAbSumm, a novel dataset designed for
aspect-based summarization of legal case decisions, sourced from the European
Court of Human Rights jurisdiction. We evaluate several abstractive
summarization models tailored for longer documents on LexAbSumm, revealing a
challenge in conditioning these models to produce aspect-specific summaries. We
release LexAbSum to facilitate research in aspect-based summarization for legal
domain. | cs.CL | LexAbSumm: Aspect-based Summarization of Legal Decisions |
2024-03-31T07:49:23Z | http://arxiv.org/pdf/2404.00590v1 | http://arxiv.org/abs/2404.00590v1 | T. Y. S. S Santosh, Kristina Kaiser, Matthias Grabmair | 2024-03-31T07:49:23Z | cs.IR, cs.CL | In this paper, we introduce CuSINeS, a negative sampling approach to enhance
the performance of Statutory Article Retrieval (SAR). CuSINeS offers three key
contributions. Firstly, it employs a curriculum-based negative sampling
strategy guiding the model to focus on easier negatives initially and
progressively tackle more difficult ones. Secondly, it leverages the
hierarchical and sequential information derived from the structural
organization of statutes to evaluate the difficulty of samples. Lastly, it
introduces a dynamic semantic difficulty assessment using the being-trained
model itself, surpassing conventional static methods like BM25, adapting the
negatives to the model's evolving competence. Experimental results on a
real-world expert-annotated SAR dataset validate the effectiveness of CuSINeS
across four different baselines, demonstrating its versatility. | cs.IR | CuSINeS: Curriculum-driven Structure Induced Negative Sampling for Statutory Article Retrieval |
2024-03-31T07:38:39Z | http://arxiv.org/pdf/2404.00589v2 | http://arxiv.org/abs/2404.00589v2 | Zhenyu Qian, Yiming Qian, Yuting Song, Fei Gao, Hai Jin, Chen Yu, Xia Xie | 2024-04-12T14:30:10Z | cs.LG, cs.CL | Handling graph data is one of the most difficult tasks. Traditional
techniques, such as those based on geometry and matrix factorization, rely on
assumptions about the data relations that become inadequate when handling large
and complex graph data. On the other hand, deep learning approaches demonstrate
promising results in handling large graph data, but they often fall short of
providing interpretable explanations. To equip the graph processing with both
high accuracy and explainability, we introduce a novel approach that harnesses
the power of a large language model (LLM), enhanced by an uncertainty-aware
module to provide a confidence score on the generated answer. We experiment
with our approach on two graph processing tasks: few-shot knowledge graph
completion and graph classification. Our results demonstrate that through
parameter efficient fine-tuning, the LLM surpasses state-of-the-art algorithms
by a substantial margin across ten diverse benchmark datasets. Moreover, to
address the challenge of explainability, we propose an uncertainty estimation
based on perturbation, along with a calibration scheme to quantify the
confidence scores of the generated answers. Our confidence measure achieves an
AUC of 0.8 or higher on seven out of the ten datasets in predicting the
correctness of the answer generated by LLM. | cs.LG | Harnessing the Power of Large Language Model for Uncertainty Aware Graph Processing |
2024-03-31T07:11:48Z | http://arxiv.org/pdf/2404.01343v2 | http://arxiv.org/abs/2404.01343v2 | Jingzhe Shi, Jialuo Li, Qinwei Ma, Zaiwen Yang, Huan Ma, Lei Li | 2024-04-15T06:03:56Z | cs.CL, cs.AI | Businesses and software platforms are increasingly turning to Large Language
Models (LLMs) such as GPT-3.5, GPT-4, GLM-3, and LLaMa-2 for chat assistance
with file access or as reasoning agents for customer service. However, current
LLM-based customer service models have limited integration with customer
profiles and lack the operational capabilities necessary for effective service.
Moreover, existing API integrations emphasize diversity over the precision and
error avoidance essential in real-world customer service scenarios. To address
these issues, we propose an LLM agent named CHOPS (CHat with custOmer Profile
in existing System), designed to: (1) efficiently utilize existing databases or
systems for accessing user information or interacting with these systems
following existing guidelines; (2) provide accurate and reasonable responses or
carry out required operations in the system while avoiding harmful operations;
and (3) leverage a combination of small and large LLMs to achieve satisfying
performance at a reasonable inference cost. We introduce a practical dataset,
the CPHOS-dataset, which includes a database, guiding files, and QA pairs
collected from CPHOS, an online platform that facilitates the organization of
simulated Physics Olympiads for high school teachers and students. We have
conducted extensive experiments to validate the performance of our proposed
CHOPS architecture using the CPHOS-dataset, with the aim of demonstrating how
LLMs can enhance or serve as alternatives to human customer service. Code for
our proposed architecture and dataset can be found at
{https://github.com/JingzheShi/CHOPS}. | cs.CL | CHOPS: CHat with custOmer Profile Systems for Customer Service with LLMs |
2024-03-31T06:28:15Z | http://arxiv.org/pdf/2404.01342v1 | http://arxiv.org/abs/2404.01342v1 | Lirui Zhao, Yue Yang, Kaipeng Zhang, Wenqi Shao, Yuxin Zhang, Yu Qiao, Ping Luo, Rongrong Ji | 2024-03-31T06:28:15Z | cs.CL, cs.AI | Text-to-image (T2I) generative models have attracted significant attention
and found extensive applications within and beyond academic research. For
example, the Civitai community, a platform for T2I innovation, currently hosts
an impressive array of 74,492 distinct models. However, this diversity presents
a formidable challenge in selecting the most appropriate model and parameters,
a process that typically requires numerous trials. Drawing inspiration from the
tool usage research of large language models (LLMs), we introduce DiffAgent, an
LLM agent designed to screen the accurate selection in seconds via API calls.
DiffAgent leverages a novel two-stage training framework, SFTA, enabling it to
accurately align T2I API responses with user input in accordance with human
preferences. To train and evaluate DiffAgent's capabilities, we present
DABench, a comprehensive dataset encompassing an extensive range of T2I APIs
from the community. Our evaluations reveal that DiffAgent not only excels in
identifying the appropriate T2I API but also underscores the effectiveness of
the SFTA training framework. Codes are available at
https://github.com/OpenGVLab/DiffAgent. | cs.CL | DiffAgent: Fast and Accurate Text-to-Image API Selection with Large Language Model |
2024-03-31T06:03:54Z | http://arxiv.org/pdf/2404.00571v1 | http://arxiv.org/abs/2404.00571v1 | Seonjeong Hwang, Yunsu Kim, Gary Geunbae Lee | 2024-03-31T06:03:54Z | cs.CL | In response to the increasing use of interactive artificial intelligence, the
demand for the capacity to handle complex questions has increased. Multi-hop
question generation aims to generate complex questions that requires multi-step
reasoning over several documents. Previous studies have predominantly utilized
end-to-end models, wherein questions are decoded based on the representation of
context documents. However, these approaches lack the ability to explain the
reasoning process behind the generated multi-hop questions. Additionally, the
question rewriting approach, which incrementally increases the question
complexity, also has limitations due to the requirement of labeling data for
intermediate-stage questions. In this paper, we introduce an end-to-end
question rewriting model that increases question complexity through sequential
rewriting. The proposed model has the advantage of training with only the final
multi-hop questions, without intermediate questions. Experimental results
demonstrate the effectiveness of our model in generating complex questions,
particularly 3- and 4-hop questions, which are appropriately paired with input
answers. We also prove that our model logically and incrementally increases the
complexity of questions, and the generated multi-hop questions are also
beneficial for training question answering models. | cs.CL | Explainable Multi-hop Question Generation: An End-to-End Approach without Intermediate Question Labeling |
2024-03-31T05:56:15Z | http://arxiv.org/pdf/2404.00570v1 | http://arxiv.org/abs/2404.00570v1 | Xingxuan Li, Xuan-Phi Nguyen, Shafiq Joty, Lidong Bing | 2024-03-31T05:56:15Z | cs.CL | Large language models (LLMs) have become the norm in natural language
processing (NLP), excelling in few-shot in-context learning (ICL) with their
remarkable abilities. Nonetheless, the success of ICL largely hinges on the
choice of few-shot demonstration examples, making the selection process
increasingly crucial. Existing methods have delved into optimizing the quantity
and semantic similarity of these examples to improve ICL performances. However,
our preliminary experiments indicate that the effectiveness of ICL is limited
by the length of the input context. Moreover, varying combinations of few-shot
demonstration examples can significantly boost accuracy across different test
samples. To address this, we propose a novel method named parallel in-context
learning (ParaICL) that effectively utilizes all demonstration examples without
exceeding the manageable input context length. ParaICL employs parallel
batching to distribute demonstration examples into different batches according
to the semantic similarities of the questions in the demonstrations to the test
question. It then computes normalized batch semantic scores for each batch. A
weighted average semantic objective, constrained by adaptive plausibility, is
applied to select the most appropriate tokens. Through extensive experiments,
we validate the effectiveness of ParaICL and conduct ablation studies to
underscore its design rationale. We further demonstrate that ParaICL can
seamlessly integrate with existing methods. | cs.CL | ParaICL: Towards Robust Parallel In-Context Learning |
2024-03-31T05:38:08Z | http://arxiv.org/pdf/2404.00569v1 | http://arxiv.org/abs/2404.00569v1 | Xiang Li, Fan Bu, Ambuj Mehrish, Yingting Li, Jiale Han, Bo Cheng, Soujanya Poria | 2024-03-31T05:38:08Z | cs.SD, cs.CL, eess.AS | Neural Text-to-Speech (TTS) systems find broad applications in voice
assistants, e-learning, and audiobook creation. The pursuit of modern models,
like Diffusion Models (DMs), holds promise for achieving high-fidelity,
real-time speech synthesis. Yet, the efficiency of multi-step sampling in
Diffusion Models presents challenges. Efforts have been made to integrate GANs
with DMs, speeding up inference by approximating denoising distributions, but
this introduces issues with model convergence due to adversarial training. To
overcome this, we introduce CM-TTS, a novel architecture grounded in
consistency models (CMs). Drawing inspiration from continuous-time diffusion
models, CM-TTS achieves top-quality speech synthesis in fewer steps without
adversarial training or pre-trained model dependencies. We further design
weighted samplers to incorporate different sampling positions into model
training with dynamic probabilities, ensuring unbiased learning throughout the
entire training process. We present a real-time mel-spectrogram generation
consistency model, validated through comprehensive evaluations. Experimental
results underscore CM-TTS's superiority over existing single-step speech
synthesis systems, representing a significant advancement in the field. | cs.SD | CM-TTS: Enhancing Real Time Text-to-Speech Synthesis Efficiency through Weighted Samplers and Consistency Models |
2024-03-31T05:20:53Z | http://arxiv.org/pdf/2404.00566v3 | http://arxiv.org/abs/2404.00566v3 | Yiqing Xie, Alex Xie, Divyanshu Sheth, Pengfei Liu, Daniel Fried, Carolyn Rose | 2024-05-08T03:14:48Z | cs.SE, cs.CL | To facilitate evaluation of code generation systems across diverse scenarios,
we present CodeBenchGen, a framework to create scalable execution-based
benchmarks that only requires light guidance from humans. Specifically, we
leverage a large language model (LLM) to convert an arbitrary piece of code
into an evaluation example, including test cases for execution-based
evaluation. We illustrate the usefulness of our framework by creating a
dataset, Exec-CSN, which includes 1,931 examples involving 293 libraries
revised from code in 367 GitHub repositories taken from the CodeSearchNet
dataset. To demonstrate the complexity and solvability of examples in Exec-CSN,
we present a human study demonstrating that 81.3% of the examples can be solved
by humans and 61% are rated as "requires effort to solve". We conduct code
generation experiments on open-source and proprietary models and analyze the
performance of both humans and models. We provide the code at
https://github.com/Veronicium/CodeBenchGen. | cs.SE | CodeBenchGen: Creating Scalable Execution-based Code Generation Benchmarks |
2024-03-31T05:14:38Z | http://arxiv.org/pdf/2404.00565v1 | http://arxiv.org/abs/2404.00565v1 | Saied Alshahrani, Hesham Haroon, Ali Elfilali, Mariama Njie, Jeanna Matthews | 2024-03-31T05:14:38Z | cs.CL | Wikipedia articles (content pages) are commonly used corpora in Natural
Language Processing (NLP) research, especially in low-resource languages other
than English. Yet, a few research studies have studied the three Arabic
Wikipedia editions, Arabic Wikipedia (AR), Egyptian Arabic Wikipedia (ARZ), and
Moroccan Arabic Wikipedia (ARY), and documented issues in the Egyptian Arabic
Wikipedia edition regarding the massive automatic creation of its articles
using template-based translation from English to Arabic without human
involvement, overwhelming the Egyptian Arabic Wikipedia with articles that do
not only have low-quality content but also with articles that do not represent
the Egyptian people, their culture, and their dialect. In this paper, we aim to
mitigate the problem of template translation that occurred in the Egyptian
Arabic Wikipedia by identifying these template-translated articles and their
characteristics through exploratory analysis and building automatic detection
systems. We first explore the content of the three Arabic Wikipedia editions in
terms of density, quality, and human contributions and utilize the resulting
insights to build multivariate machine learning classifiers leveraging
articles' metadata to detect the template-translated articles automatically. We
then publicly deploy and host the best-performing classifier, XGBoost, as an
online application called EGYPTIAN WIKIPEDIA SCANNER and release the extracted,
filtered, and labeled datasets to the research community to benefit from our
datasets and the online, web-based detection system. | cs.CL | Leveraging Corpus Metadata to Detect Template-based Translation: An Exploratory Case Study of the Egyptian Arabic Wikipedia Edition |
2024-03-31T04:36:57Z | http://arxiv.org/pdf/2404.00557v1 | http://arxiv.org/abs/2404.00557v1 | Weihao Zeng, Dayuan Fu, Keqing He, Yejie Wang, Yukai Xu, Weiran Xu | 2024-03-31T04:36:57Z | cs.CL | Language models pre-trained on general text have achieved impressive results
in diverse fields. Yet, the distinct linguistic characteristics of
task-oriented dialogues (TOD) compared to general text limit the practical
utility of existing language models. Current task-oriented dialogue
pre-training methods overlook the one-to-many property of conversations, where
multiple responses can be appropriate given the same conversation context. In
this paper, we propose a novel dialogue pre-training model called DivTOD, which
collaborates with LLMs to learn diverse task-oriented dialogue representations.
DivTOD guides LLMs in transferring diverse knowledge to smaller models while
removing domain knowledge that contradicts task-oriented dialogues. Experiments
show that our model outperforms strong TOD baselines on various downstream
dialogue tasks and learns the intrinsic diversity of task-oriented dialogues. | cs.CL | DivTOD: Unleashing the Power of LLMs for Diversifying Task-Oriented Dialogue Representations |
2024-03-31T02:05:40Z | http://arxiv.org/pdf/2404.00530v1 | http://arxiv.org/abs/2404.00530v1 | Hritik Bansal, Ashima Suvarna, Gantavya Bhatt, Nanyun Peng, Kai-Wei Chang, Aditya Grover | 2024-03-31T02:05:40Z | cs.CL, cs.AI, cs.LG | A common technique for aligning large language models (LLMs) relies on
acquiring human preferences by comparing multiple generations conditioned on a
fixed context. This only leverages the pairwise comparisons when the
generations are placed in an identical context. However, such conditional
rankings often fail to capture the complex and multidimensional aspects of
human preferences. In this work, we revisit the traditional paradigm of
preference acquisition and propose a new axis that is based on eliciting
preferences jointly over the instruction-response pairs. While prior preference
optimizations are designed for conditional ranking protocols (e.g., DPO), our
proposed preference acquisition protocol introduces DOVE, a new preference
optimization objective that upweights the joint probability of the chosen
instruction-response pair over the rejected instruction-response pair.
Interestingly, we find that the LLM trained with joint instruction-response
preference data using DOVE outperforms the LLM trained with DPO by 5.2% and
3.3% win-rate for the summarization and open-ended dialogue datasets,
respectively. Our findings reveal that joint preferences over instruction and
response pairs can significantly enhance the alignment of LLMs by tapping into
a broader spectrum of human preference elicitation. The data and code is
available at https://github.com/Hritikbansal/dove. | cs.CL | Comparing Bad Apples to Good Oranges: Aligning Large Language Models via Joint Preference Optimization |
2024-03-31T01:16:02Z | http://arxiv.org/pdf/2404.00511v3 | http://arxiv.org/abs/2404.00511v3 | Zebang Cheng, Fuqiang Niu, Yuxiang Lin, Zhi-Qi Cheng, Bowen Zhang, Xiaojiang Peng | 2024-04-11T05:14:35Z | cs.CL, cs.CV, cs.MM | This paper presents our winning submission to Subtask 2 of SemEval 2024 Task
3 on multimodal emotion cause analysis in conversations. We propose a novel
Multimodal Emotion Recognition and Multimodal Emotion Cause Extraction
(MER-MCE) framework that integrates text, audio, and visual modalities using
specialized emotion encoders. Our approach sets itself apart from
top-performing teams by leveraging modality-specific features for enhanced
emotion understanding and causality inference. Experimental evaluation
demonstrates the advantages of our multimodal approach, with our submission
achieving a competitive weighted F1 score of 0.3435, ranking third with a
margin of only 0.0339 behind the 1st team and 0.0025 behind the 2nd team.
Project: https://github.com/MIPS-COLT/MER-MCE.git | cs.CL | MIPS at SemEval-2024 Task 3: Multimodal Emotion-Cause Pair Extraction in Conversations with Multimodal Language Models |
2024-03-31T00:38:02Z | http://arxiv.org/pdf/2404.01339v1 | http://arxiv.org/abs/2404.01339v1 | Rohan Chaudhury, Mihir Godbole, Aakash Garg, Jinsil Hwaryoung Seo | 2024-03-31T00:38:02Z | cs.CL, cs.AI, cs.HC | Contemporary conversational systems often present a significant limitation:
their responses lack the emotional depth and disfluent characteristic of human
interactions. This absence becomes particularly noticeable when users seek more
personalized and empathetic interactions. Consequently, this makes them seem
mechanical and less relatable to human users. Recognizing this gap, we embarked
on a journey to humanize machine communication, to ensure AI systems not only
comprehend but also resonate. To address this shortcoming, we have designed an
innovative speech synthesis pipeline. Within this framework, a cutting-edge
language model introduces both human-like emotion and disfluencies in a
zero-shot setting. These intricacies are seamlessly integrated into the
generated text by the language model during text generation, allowing the
system to mirror human speech patterns better, promoting more intuitive and
natural user interactions. These generated elements are then adeptly
transformed into corresponding speech patterns and emotive sounds using a
rule-based approach during the text-to-speech phase. Based on our experiments,
our novel system produces synthesized speech that's almost indistinguishable
from genuine human communication, making each interaction feel more personal
and authentic. | cs.CL | Humane Speech Synthesis through Zero-Shot Emotion and Disfluency Generation |
2024-03-30T23:51:25Z | http://arxiv.org/pdf/2404.00500v1 | http://arxiv.org/abs/2404.00500v1 | Ondřej Draganov, Steven Skiena | 2024-03-30T23:51:25Z | cs.CL, math.AT | Word embeddings represent language vocabularies as clouds of $d$-dimensional
points. We investigate how information is conveyed by the general shape of
these clouds, outside of representing the semantic meaning of each token.
Specifically, we use the notion of persistent homology from topological data
analysis (TDA) to measure the distances between language pairs from the shape
of their unlabeled embeddings. We use these distance matrices to construct
language phylogenetic trees over 81 Indo-European languages. Careful evaluation
shows that our reconstructed trees exhibit strong similarities to the reference
tree. | cs.CL | The Shape of Word Embeddings: Recognizing Language Phylogenies through Topological Data Analysis |
2024-03-30T23:28:05Z | http://arxiv.org/pdf/2404.00495v1 | http://arxiv.org/abs/2404.00495v1 | Victor Gallego | 2024-03-30T23:28:05Z | cs.CL, cs.AI | State-of-the-art language model fine-tuning techniques, such as Direct
Preference Optimization (DPO), restrict user control by hard-coding predefined
behaviors into the model. To address this, we propose a novel method,
Configurable Safety Tuning (CST), that augments DPO using synthetic preference
data to facilitate flexible safety configuration of LLMs at inference time. CST
overcomes the constraints of vanilla DPO by introducing a system prompt
specifying safety configurations, enabling LLM deployers to disable/enable
safety preferences based on their need, just changing the system prompt. Our
experimental evaluations indicate that CST successfully manages different
safety configurations and retains the original functionality of LLMs, showing
it is a robust method for configurable deployment. Data and models available at
https://github.com/vicgalle/configurable-safety-tuning | cs.CL | Configurable Safety Tuning of Language Models with Synthetic Preference Data |
2024-03-30T23:22:51Z | http://arxiv.org/pdf/2404.00492v1 | http://arxiv.org/abs/2404.00492v1 | Keyuan Cheng, Gang Lin, Haoyang Fei, Yuxuan zhai, Lu Yu, Muhammad Asif Ali, Lijie Hu, Di Wang | 2024-03-30T23:22:51Z | cs.CL, cs.AI, cs.LG | Multi-hop question answering (MQA) under knowledge editing (KE) has garnered
significant attention in the era of large language models. However, existing
models for MQA under KE exhibit poor performance when dealing with questions
containing explicit temporal contexts. To address this limitation, we propose a
novel framework, namely TEMPoral knowLEdge augmented Multi-hop Question
Answering (TEMPLE-MQA). Unlike previous methods, TEMPLE-MQA first constructs a
time-aware graph (TAG) to store edit knowledge in a structured manner. Then,
through our proposed inference path, structural retrieval, and joint reasoning
stages, TEMPLE-MQA effectively discerns temporal contexts within the question
query. Experiments on benchmark datasets demonstrate that TEMPLE-MQA
significantly outperforms baseline models. Additionally, we contribute a new
dataset, namely TKEMQA, which serves as the inaugural benchmark tailored
specifically for MQA with temporal scopes. | cs.CL | Multi-hop Question Answering under Temporal Knowledge Editing |
2024-03-30T23:07:58Z | http://arxiv.org/pdf/2404.00489v1 | http://arxiv.org/abs/2404.00489v1 | Muhammad Asif Ali, Zhengping Li, Shu Yang, Keyuan Cheng, Yang Cao, Tianhao Huang, Lijie Hu, Lu Yu, Di Wang | 2024-03-30T23:07:58Z | cs.CL, cs.AI, cs.LG | Large language models (LLMs) have shown exceptional abilities for multiple
different natural language processing tasks. While prompting is a crucial tool
for LLM inference, we observe that there is a significant cost associated with
exceedingly lengthy prompts. Existing attempts to compress lengthy prompts lead
to sub-standard results in terms of readability and interpretability of the
compressed prompt, with a detrimental impact on prompt utility. To address
this, we propose PROMPT-SAW: Prompt compresSion via Relation AWare graphs, an
effective strategy for prompt compression over task-agnostic and task-aware
prompts. PROMPT-SAW uses the prompt's textual information to build a graph,
later extracts key information elements in the graph to come up with the
compressed prompt. We also propose GSM8K-AUG, i.e., an extended version of the
existing GSM8k benchmark for task-agnostic prompts in order to provide a
comprehensive evaluation platform. Experimental evaluation using benchmark
datasets shows that prompts compressed by PROMPT-SAW are not only better in
terms of readability, but they also outperform the best-performing baseline
models by up to 14.3 and 13.7 respectively for task-aware and task-agnostic
settings while compressing the original prompt text by 33.0 and 56.7. | cs.CL | PROMPT-SAW: Leveraging Relation-Aware Graphs for Textual Prompt Compression |
2024-03-30T23:06:34Z | http://arxiv.org/pdf/2404.00488v1 | http://arxiv.org/abs/2404.00488v1 | Ritesh Sarkhel, Xiaoqi Ren, Lauro Beltrao Costa, Guolong Su, Vincent Perot, Yanan Xie, Emmanouil Koukoumidis, Arnab Nandi | 2024-03-30T23:06:34Z | cs.CL, cs.AI, cs.LG | A visually rich document (VRD) utilizes visual features along with linguistic
cues to disseminate information. Training a custom extractor that identifies
named entities from a document requires a large number of instances of the
target document type annotated at textual and visual modalities. This is an
expensive bottleneck in enterprise scenarios, where we want to train custom
extractors for thousands of different document types in a scalable way.
Pre-training an extractor model on unlabeled instances of the target document
type, followed by a fine-tuning step on human-labeled instances does not work
in these scenarios, as it surpasses the maximum allowable training time
allocated for the extractor. We address this scenario by proposing a
Noise-Aware Training method or NAT in this paper. Instead of acquiring
expensive human-labeled documents, NAT utilizes weakly labeled documents to
train an extractor in a scalable way. To avoid degradation in the model's
quality due to noisy, weakly labeled samples, NAT estimates the confidence of
each training sample and incorporates it as uncertainty measure during
training. We train multiple state-of-the-art extractor models using NAT.
Experiments on a number of publicly available and in-house datasets show that
NAT-trained models are not only robust in performance -- it outperforms a
transfer-learning baseline by up to 6% in terms of macro-F1 score, but it is
also more label-efficient -- it reduces the amount of human-effort required to
obtain comparable performance by up to 73%. | cs.CL | Noise-Aware Training of Layout-Aware Language Models |
2024-03-30T22:41:05Z | http://arxiv.org/pdf/2404.00486v1 | http://arxiv.org/abs/2404.00486v1 | Shu Yang, Jiayuan Su, Han Jiang, Mengdi Li, Keyuan Cheng, Muhammad Asif Ali, Lijie Hu, Di Wang | 2024-03-30T22:41:05Z | cs.CL, cs.AI | With the rise of large language models (LLMs), ensuring they embody the
principles of being helpful, honest, and harmless (3H), known as Human
Alignment, becomes crucial. While existing alignment methods like RLHF, DPO,
etc., effectively fine-tune LLMs to match preferences in the preference
dataset, they often lead LLMs to highly receptive human input and external
evidence, even when this information is poisoned. This leads to a tendency for
LLMs to be Adaptive Chameleons when external evidence conflicts with their
parametric memory. This exacerbates the risk of LLM being attacked by external
poisoned data, which poses a significant security risk to LLM system
applications such as Retrieval-augmented generation (RAG). To address the
challenge, we propose a novel framework: Dialectical Alignment (DA), which (1)
utilizes AI feedback to identify optimal strategies for LLMs to navigate
inter-context conflicts and context-memory conflicts with different external
evidence in context window (i.e., different ratios of poisoned factual
contexts); (2) constructs the SFT dataset as well as the preference dataset
based on the AI feedback and strategies above; (3) uses the above datasets for
LLM alignment to defense poisoned context attack while preserving the
effectiveness of in-context knowledge editing. Our experiments show that the
dialectical alignment model improves poisoned data attack defense by 20 and
does not require any additional prompt engineering or prior declaration of
``you may be attacked`` to the LLMs' context window. | cs.CL | Dialectical Alignment: Resolving the Tension of 3H and Security Threats of LLMs |
2024-03-30T22:27:21Z | http://arxiv.org/pdf/2404.00484v1 | http://arxiv.org/abs/2404.00484v1 | Aryo Pradipta Gema, Giwon Hong, Pasquale Minervini, Luke Daines, Beatrice Alex | 2024-03-30T22:27:21Z | cs.CL | The NLI4CT task assesses Natural Language Inference systems in predicting
whether hypotheses entail or contradict evidence from Clinical Trial Reports.
In this study, we evaluate various Large Language Models (LLMs) with multiple
strategies, including Chain-of-Thought, In-Context Learning, and
Parameter-Efficient Fine-Tuning (PEFT). We propose a PEFT method to improve the
consistency of LLMs by merging adapters that were fine-tuned separately using
triplet and language modelling objectives. We found that merging the two PEFT
adapters improves the F1 score (+0.0346) and consistency (+0.152) of the LLMs.
However, our novel methods did not produce more accurate results than GPT-4 in
terms of faithfulness and consistency. Averaging the three metrics, GPT-4 ranks
joint-first in the competition with 0.8328. Finally, our contamination analysis
with GPT-4 indicates that there was no test data leakage. | cs.CL | Edinburgh Clinical NLP at SemEval-2024 Task 2: Fine-tune your model unless you have access to GPT-4 |
2024-03-30T22:20:08Z | http://arxiv.org/pdf/2404.00482v2 | http://arxiv.org/abs/2404.00482v2 | Jakub Piskorski, Michał Marcińczuk, Roman Yangarber | 2024-04-07T16:56:35Z | cs.CL, cs.AI, cs.LG | This paper presents a corpus manually annotated with named entities for six
Slavic languages - Bulgarian, Czech, Polish, Slovenian, Russian, and Ukrainian.
This work is the result of a series of shared tasks, conducted in 2017-2023 as
a part of the Workshops on Slavic Natural Language Processing. The corpus
consists of 5 017 documents on seven topics. The documents are annotated with
five classes of named entities. Each entity is described by a category, a
lemma, and a unique cross-lingual identifier. We provide two train-tune dataset
splits - single topic out and cross topics. For each split, we set benchmarks
using a transformer-based neural network architecture with the pre-trained
multilingual models - XLM-RoBERTa-large for named entity mention recognition
and categorization, and mT5-large for named entity lemmatization and linking. | cs.CL | Cross-lingual Named Entity Corpus for Slavic Languages |
2024-03-30T20:47:55Z | http://arxiv.org/pdf/2404.00474v2 | http://arxiv.org/abs/2404.00474v2 | Neil Band, Xuechen Li, Tengyu Ma, Tatsunori Hashimoto | 2024-06-04T22:39:58Z | cs.LG, cs.AI, cs.CL, stat.ML | Language models (LMs) may lead their users to make suboptimal downstream
decisions when they confidently hallucinate. This issue can be mitigated by
having the LM verbally convey the probability that its claims are correct, but
existing models cannot produce long-form text with calibrated confidence
statements. Through the lens of decision-making, we define linguistic
calibration for long-form generations: an LM is linguistically calibrated if
its generations enable its users to make calibrated probabilistic predictions.
This definition enables a training framework where a supervised finetuning step
bootstraps an LM to emit long-form generations with confidence statements such
as "I estimate a 30% chance of..." or "I am certain that...", followed by a
reinforcement learning step which rewards generations that enable a user to
provide calibrated answers to related questions. We linguistically calibrate
Llama 2 7B and find in automated and human evaluations of long-form generations
that it is significantly more calibrated than strong finetuned factuality
baselines with comparable accuracy. These findings generalize under significant
domain shifts to scientific and biomedical questions and to an entirely
held-out person biography generation task. Our results demonstrate that
long-form generations may be calibrated end-to-end by constructing an objective
in the space of the predictions that users make in downstream decision-making. | cs.LG | Linguistic Calibration of Long-Form Generations |
2024-03-30T20:05:41Z | http://arxiv.org/pdf/2404.00463v1 | http://arxiv.org/abs/2404.00463v1 | Hannah Chen, Yangfeng Ji, David Evans | 2024-03-30T20:05:41Z | cs.CL, cs.CY, cs.LG | Statistical fairness stipulates equivalent outcomes for every protected
group, whereas causal fairness prescribes that a model makes the same
prediction for an individual regardless of their protected characteristics.
Counterfactual data augmentation (CDA) is effective for reducing bias in NLP
models, yet models trained with CDA are often evaluated only on metrics that
are closely tied to the causal fairness notion; similarly, sampling-based
methods designed to promote statistical fairness are rarely evaluated for
causal fairness. In this work, we evaluate both statistical and causal
debiasing methods for gender bias in NLP models, and find that while such
methods are effective at reducing bias as measured by the targeted metric, they
do not necessarily improve results on other bias metrics. We demonstrate that
combinations of statistical and causal debiasing techniques are able to reduce
bias measured through both types of metrics. | cs.CL | Addressing Both Statistical and Causal Gender Fairness in NLP Models |
2024-03-30T20:02:36Z | http://arxiv.org/pdf/2404.00461v1 | http://arxiv.org/abs/2404.00461v1 | Xiaopeng Xie, Ming Yan, Xiwen Zhou, Chenlong Zhao, Suli Wang, Yong Zhang, Joey Tianyi Zhou | 2024-03-30T20:02:36Z | cs.LG, cs.AI, cs.CL, cs.CR, 68T50, I.2.7 | Prompt-based learning paradigm has demonstrated remarkable efficacy in
enhancing the adaptability of pretrained language models (PLMs), particularly
in few-shot scenarios. However, this learning paradigm has been shown to be
vulnerable to backdoor attacks. The current clean-label attack, employing a
specific prompt as a trigger, can achieve success without the need for external
triggers and ensure correct labeling of poisoned samples, which is more
stealthy compared to the poisoned-label attack, but on the other hand, it faces
significant issues with false activations and poses greater challenges,
necessitating a higher rate of poisoning. Using conventional negative data
augmentation methods, we discovered that it is challenging to trade off between
effectiveness and stealthiness in a clean-label setting. In addressing this
issue, we are inspired by the notion that a backdoor acts as a shortcut and
posit that this shortcut stems from the contrast between the trigger and the
data utilized for poisoning. In this study, we propose a method named
Contrastive Shortcut Injection (CSI), by leveraging activation values,
integrates trigger design and data selection strategies to craft stronger
shortcut features. With extensive experiments on full-shot and few-shot text
classification tasks, we empirically validate CSI's high effectiveness and high
stealthiness at low poisoning rates. Notably, we found that the two approaches
play leading roles in full-shot and few-shot settings, respectively. | cs.LG | Shortcuts Arising from Contrast: Effective and Covert Clean-Label Attacks in Prompt-Based Learning |
2024-03-30T19:46:59Z | http://arxiv.org/pdf/2404.00459v1 | http://arxiv.org/abs/2404.00459v1 | Eli Schwartz, Leshem Choshen, Joseph Shtok, Sivan Doveh, Leonid Karlinsky, Assaf Arbelle | 2024-03-30T19:46:59Z | cs.CL | Language models struggle with handling numerical data and performing
arithmetic operations. We hypothesize that this limitation can be partially
attributed to non-intuitive textual numbers representation. When a digit is
read or generated by a causal language model it does not know its place value
(e.g. thousands vs. hundreds) until the entire number is processed. To address
this issue, we propose a simple adjustment to how numbers are represented by
including the count of digits before each number. For instance, instead of
"42", we suggest using "{2:42}" as the new format. This approach, which we term
NumeroLogic, offers an added advantage in number generation by serving as a
Chain of Thought (CoT). By requiring the model to consider the number of digits
first, it enhances the reasoning process before generating the actual number.
We use arithmetic tasks to demonstrate the effectiveness of the NumeroLogic
formatting. We further demonstrate NumeroLogic applicability to general natural
language modeling, improving language understanding performance in the MMLU
benchmark. | cs.CL | NumeroLogic: Number Encoding for Enhanced LLMs' Numerical Reasoning |
2024-03-30T19:45:04Z | http://arxiv.org/pdf/2404.00458v1 | http://arxiv.org/abs/2404.00458v1 | Vivek Khetan | 2024-03-30T19:45:04Z | cs.CL, cs.IR | This position paper proposes a systematic approach towards developing a
framework to help select the most effective embedding models for natural
language processing (NLP) tasks, addressing the challenge posed by the
proliferation of both proprietary and open-source encoder models. | cs.CL | Beyond One-Size-Fits-All: Multi-Domain, Multi-Task Framework for Embedding Model Selection |
2024-03-30T19:43:45Z | http://arxiv.org/pdf/2404.00457v1 | http://arxiv.org/abs/2404.00457v1 | Letian Peng, Zilong Wang, Feng Yao, Zihan Wang, Jingbo Shang | 2024-03-30T19:43:45Z | cs.CL | Information extraction (IE) is a fundamental area in natural language
processing where prompting large language models (LLMs), even with in-context
examples, cannot defeat small LMs tuned on very small IE datasets. We observe
that IE tasks, such as named entity recognition and relation extraction, all
focus on extracting important information, which can be formalized as a
label-to-span matching. In this paper, we propose a novel framework MetaIE to
build a small LM as meta-model by learning to extract "important information",
i.e., the meta-understanding of IE, so that this meta-model can be adapted to
all kind of IE tasks effectively and efficiently. Specifically, MetaIE obtains
the small LM via a symbolic distillation from an LLM following the
label-to-span scheme. We construct the distillation dataset via sampling
sentences from language model pre-training datasets (e.g., OpenWebText in our
implementation) and prompting an LLM to identify the typed spans of "important
information". We evaluate the meta-model under the few-shot adaptation setting.
Extensive results on 13 datasets from 6 IE tasks confirm that MetaIE can offer
a better starting point for few-shot tuning on IE datasets and outperform other
meta-models from (1) vanilla language model pre-training, (2) multi-IE-task
pre-training with human annotations, and (3) single-IE-task symbolic
distillation from LLM. Moreover, we provide comprehensive analyses of MetaIE,
such as the size of the distillation dataset, the meta-model architecture, and
the size of the meta-model. | cs.CL | MetaIE: Distilling a Meta Model from LLM for All Kinds of Information Extraction Tasks |
2024-03-30T18:41:51Z | http://arxiv.org/pdf/2404.00450v2 | http://arxiv.org/abs/2404.00450v2 | Tenghao Huang, Dongwon Jung, Muhao Chen | 2024-04-04T05:33:07Z | cs.CL, cs.AI, cs.IR, cs.LG | Recent advancements in integrating external tools with Large Language Models
(LLMs) have opened new frontiers, with applications in mathematical reasoning,
code generators, and smart assistants. However, existing methods, relying on
simple one-time retrieval strategies, fall short on effectively and accurately
shortlisting relevant tools. This paper introduces a novel PLUTO (Planning,
Learning, and Understanding for TOols) approach, encompassing
`Plan-and-Retrieve (P&R)` and `Edit-and-Ground (E&G)` paradigms. The P&R
paradigm consists of a neural retrieval module for shortlisting relevant tools
and an LLM-based query planner that decomposes complex queries into actionable
tasks, enhancing the effectiveness of tool utilization. The E&G paradigm
utilizes LLMs to enrich tool descriptions based on user scenarios, bridging the
gap between user queries and tool functionalities. Experiment results
demonstrate that these paradigms significantly improve the recall and NDCG in
tool retrieval tasks, significantly surpassing current state-of-the-art models. | cs.CL | Planning and Editing What You Retrieve for Enhanced Tool Learning |
2024-03-30T18:11:39Z | http://arxiv.org/pdf/2404.00439v1 | http://arxiv.org/abs/2404.00439v1 | Alex Nguyen, Zilong Wang, Jingbo Shang, Dheeraj Mekala | 2024-03-30T18:11:39Z | cs.CL | The application of natural language processing models to PDF documents is
pivotal for various business applications yet the challenge of training models
for this purpose persists in businesses due to specific hurdles. These include
the complexity of working with PDF formats that necessitate parsing text and
layout information for curating training data and the lack of
privacy-preserving annotation tools. This paper introduces DOCMASTER, a unified
platform designed for annotating PDF documents, model training, and inference,
tailored to document question-answering. The annotation interface enables users
to input questions and highlight text spans within the PDF file as answers,
saving layout information and text spans accordingly. Furthermore, DOCMASTER
supports both state-of-the-art layout-aware and text models for comprehensive
training purposes. Importantly, as annotations, training, and inference occur
on-device, it also safeguards privacy. The platform has been instrumental in
driving several research prototypes concerning document analysis such as the AI
assistant utilized by University of California San Diego's (UCSD) International
Services and Engagement Office (ISEO) for processing a substantial volume of
PDF documents. | cs.CL | DOCMASTER: A Unified Platform for Annotation, Training, & Inference in Document Question-Answering |
2024-03-30T17:59:43Z | http://arxiv.org/pdf/2404.00437v1 | http://arxiv.org/abs/2404.00437v1 | Jaime González-González, Francisco de Arriba-Pérez, Silvia García-Méndez, Andrea Busto-Castiñeira, Francisco J. González-Castaño | 2024-03-30T17:59:43Z | cs.CL, cs.AI, cs.LG | Automatic legal text classification systems have been proposed in the
literature to address knowledge extraction from judgments and detect their
aspects. However, most of these systems are black boxes even when their models
are interpretable. This may raise concerns about their trustworthiness.
Accordingly, this work contributes with a system combining Natural Language
Processing (NLP) with Machine Learning (ML) to classify legal texts in an
explainable manner. We analyze the features involved in the decision and the
threshold bifurcation values of the decision paths of tree structures and
present this information to the users in natural language. This is the first
work on automatic analysis of legal texts combining NLP and ML along with
Explainable Artificial Intelligence techniques to automatically make the
models' decisions understandable to end users. Furthermore, legal experts have
validated our solution, and this knowledge has also been incorporated into the
explanation process as "expert-in-the-loop" dictionaries. Experimental results
on an annotated data set in law categories by jurisdiction demonstrate that our
system yields competitive classification performance, with accuracy values well
above 90%, and that its automatic explanations are easily understandable even
to non-expert users. | cs.CL | Automatic explanation of the classification of Spanish legal judgments in jurisdiction-dependent law categories with tree estimators |
2024-03-30T17:49:34Z | http://arxiv.org/pdf/2404.01338v1 | http://arxiv.org/abs/2404.01338v1 | Silvia García-Méndez, Francisco de Arriba-Pérez, Ana Barros-Vila, Francisco J. González-Castaño, Enrique Costa-Montenegro | 2024-03-30T17:49:34Z | cs.CL, cs.CE, cs.IR, cs.LG, q-fin.ST | Financial news items are unstructured sources of information that can be
mined to extract knowledge for market screening applications. Manual extraction
of relevant information from the continuous stream of finance-related news is
cumbersome and beyond the skills of many investors, who, at most, can follow a
few sources and authors. Accordingly, we focus on the analysis of financial
news to identify relevant text and, within that text, forecasts and
predictions. We propose a novel Natural Language Processing (NLP) system to
assist investors in the detection of relevant financial events in unstructured
textual sources by considering both relevance and temporality at the discursive
level. Firstly, we segment the text to group together closely related text.
Secondly, we apply co-reference resolution to discover internal dependencies
within segments. Finally, we perform relevant topic modelling with Latent
Dirichlet Allocation (LDA) to separate relevant from less relevant text and
then analyse the relevant text using a Machine Learning-oriented temporal
approach to identify predictions and speculative statements. We created an
experimental data set composed of 2,158 financial news items that were manually
labelled by NLP researchers to evaluate our solution. The ROUGE-L values for
the identification of relevant text and predictions/forecasts were 0.662 and
0.982, respectively. To our knowledge, this is the first work to jointly
consider relevance and temporality at the discursive level. It contributes to
the transfer of human associative discourse capabilities to expert systems
through the combination of multi-paragraph topic segmentation and co-reference
resolution to separate author expression patterns, topic modelling with LDA to
detect relevant text, and discursive temporality analysis to identify forecasts
and predictions within this text. | cs.CL | Automatic detection of relevant information, predictions and forecasts in financial news through topic modelling with Latent Dirichlet Allocation |
2024-03-30T16:54:45Z | http://arxiv.org/pdf/2404.00419v1 | http://arxiv.org/abs/2404.00419v1 | Sonal Kumar, Sreyan Ghosh, S Sakshi, Utkarsh Tyagi, Dinesh Manocha | 2024-03-30T16:54:45Z | cs.CV, cs.CL | Open-vocabulary vision-language models (VLMs) like CLIP, trained using
contrastive loss, have emerged as a promising new paradigm for text-to-image
retrieval. However, do VLMs understand compound nouns (CNs) (e.g., lab coat) as
well as they understand nouns (e.g., lab)? We curate Compun, a novel benchmark
with 400 unique and commonly used CNs, to evaluate the effectiveness of VLMs in
interpreting CNs. The Compun benchmark challenges a VLM for text-to-image
retrieval where, given a text prompt with a CN, the task is to select the
correct image that shows the CN among a pair of distractor images that show the
constituent nouns that make up the CN. Next, we perform an in-depth analysis to
highlight CLIPs' limited understanding of certain types of CNs. Finally, we
present an alternative framework that moves beyond hand-written templates for
text prompts widely used by CLIP-like models. We employ a Large Language Model
to generate multiple diverse captions that include the CN as an object in the
scene described by the caption. Our proposed method improves CN understanding
of CLIP by 8.25% on Compun. Code and benchmark are available at:
https://github.com/sonalkum/Compun | cs.CV | Do Vision-Language Models Understand Compound Nouns? |
2024-03-30T16:47:06Z | http://arxiv.org/pdf/2404.00415v1 | http://arxiv.org/abs/2404.00415v1 | Chandra Kiran Reddy Evuru, Sreyan Ghosh, Sonal Kumar, Ramaneswaran S, Utkarsh Tyagi, Dinesh Manocha | 2024-03-30T16:47:06Z | cs.CL | We present CoDa (Constrained Generation based Data Augmentation), a
controllable, effective, and training-free data augmentation technique for
low-resource (data-scarce) NLP. Our approach is based on prompting
off-the-shelf instruction-following Large Language Models (LLMs) for generating
text that satisfies a set of constraints. Precisely, we extract a set of simple
constraints from every instance in the low-resource dataset and verbalize them
to prompt an LLM to generate novel and diverse training instances. Our findings
reveal that synthetic data that follows simple constraints in the downstream
dataset act as highly effective augmentations, and CoDa can achieve this
without intricate decoding-time constrained generation techniques or
fine-tuning with complex algorithms that eventually make the model biased
toward the small number of training instances. Additionally, CoDa is the first
framework that provides users explicit control over the augmentation generation
process, thereby also allowing easy adaptation to several domains. We
demonstrate the effectiveness of CoDa across 11 datasets spanning 3 tasks and 3
low-resource settings. CoDa outperforms all our baselines, qualitatively and
quantitatively, with improvements of 0.12%-7.19%. Code is available here:
https://github.com/Sreyan88/CoDa | cs.CL | CoDa: Constrained Generation based Data Augmentation for Low-Resource NLP |
2024-03-30T16:46:25Z | http://arxiv.org/pdf/2404.08665v1 | http://arxiv.org/abs/2404.08665v1 | Silvia García-Méndez, Francisco de Arriba-Pérez, Ana Barros-Vila, Francisco J. González-Castaño | 2024-03-30T16:46:25Z | cs.IR, cs.CL, cs.LG, cs.SI, q-fin.TR | Microblogging platforms, of which Twitter is a representative example, are
valuable information sources for market screening and financial models. In
them, users voluntarily provide relevant information, including educated
knowledge on investments, reacting to the state of the stock markets in
real-time and, often, influencing this state. We are interested in the user
forecasts in financial, social media messages expressing opportunities and
precautions about assets. We propose a novel Targeted Aspect-Based Emotion
Analysis (TABEA) system that can individually discern the financial emotions
(positive and negative forecasts) on the different stock market assets in the
same tweet (instead of making an overall guess about that whole tweet). It is
based on Natural Language Processing (NLP) techniques and Machine Learning
streaming algorithms. The system comprises a constituency parsing module for
parsing the tweets and splitting them into simpler declarative clauses; an
offline data processing module to engineer textual, numerical and categorical
features and analyse and select them based on their relevance; and a stream
classification module to continuously process tweets on-the-fly. Experimental
results on a labelled data set endorse our solution. It achieves over 90%
precision for the target emotions, financial opportunity, and precaution on
Twitter. To the best of our knowledge, no prior work in the literature has
addressed this problem despite its practical interest in decision-making, and
we are not aware of any previous NLP nor online Machine Learning approaches to
TABEA. | cs.IR | Targeted aspect-based emotion analysis to detect opportunities and precaution in financial Twitter messages |
2024-03-30T16:40:10Z | http://arxiv.org/pdf/2404.01337v1 | http://arxiv.org/abs/2404.01337v1 | Silvia García-Méndez, Francisco de Arriba-Pérez, Ana Barros-Vila, Francisco J. González-Castaño | 2024-03-30T16:40:10Z | cs.CL, cs.CE, cs.IR, cs.LG, q-fin.ST | Finance-related news such as Bloomberg News, CNN Business and Forbes are
valuable sources of real data for market screening systems. In news, an expert
shares opinions beyond plain technical analyses that include context such as
political, sociological and cultural factors. In the same text, the expert
often discusses the performance of different assets. Some key statements are
mere descriptions of past events while others are predictions. Therefore,
understanding the temporality of the key statements in a text is essential to
separate context information from valuable predictions. We propose a novel
system to detect the temporality of finance-related news at discourse level
that combines Natural Language Processing and Machine Learning techniques, and
exploits sophisticated features such as syntactic and semantic dependencies.
More specifically, we seek to extract the dominant tenses of the main
statements, which may be either explicit or implicit. We have tested our system
on a labelled dataset of finance-related news annotated by researchers with
knowledge in the field. Experimental results reveal a high detection precision
compared to an alternative rule-based baseline approach. Ultimately, this
research contributes to the state-of-the-art of market screening by identifying
predictive knowledge for financial decision making. | cs.CL | Detection of Temporality at Discourse Level on Financial News by Combining Natural Language Processing and Machine Learning |
2024-03-30T16:14:46Z | http://arxiv.org/pdf/2404.00406v1 | http://arxiv.org/abs/2404.00406v1 | Marc Feger, Stefan Dietze | 2024-03-30T16:14:46Z | cs.CL, cs.AI | Twitter has emerged as a global hub for engaging in online conversations and
as a research corpus for various disciplines that have recognized the
significance of its user-generated content. Argument mining is an important
analytical task for processing and understanding online discourse.
Specifically, it aims to identify the structural elements of arguments, denoted
as information and inference. These elements, however, are not static and may
require context within the conversation they are in, yet there is a lack of
data and annotation frameworks addressing this dynamic aspect on Twitter. We
contribute TACO, the first dataset of Twitter Arguments utilizing 1,814 tweets
covering 200 entire conversations spanning six heterogeneous topics annotated
with an agreement of 0.718 Krippendorff's alpha among six experts. Second, we
provide our annotation framework, incorporating definitions from the Cambridge
Dictionary, to define and identify argument components on Twitter. Our
transformer-based classifier achieves an 85.06\% macro F1 baseline score in
detecting arguments. Moreover, our data reveals that Twitter users tend to
engage in discussions involving informed inferences and information. TACO
serves multiple purposes, such as training tweet classifiers to manage tweets
based on inference and information elements, while also providing valuable
insights into the conversational reply patterns of tweets. | cs.CL | TACO -- Twitter Arguments from COnversations |
2024-03-30T15:59:17Z | http://arxiv.org/pdf/2404.00403v1 | http://arxiv.org/abs/2404.00403v1 | Guimin Hu, Zhihong Zhu, Daniel Hershcovich, Hasti Seifi, Jiayuan Xie | 2024-03-30T15:59:17Z | cs.CL | Multimodal emotion recognition in conversation (MERC) and multimodal
emotion-cause pair extraction (MECPE) has recently garnered significant
attention. Emotions are the expression of affect or feelings; responses to
specific events, thoughts, or situations are known as emotion causes. Both are
like two sides of a coin, collectively describing human behaviors and intents.
However, most existing works treat MERC and MECPE as separate tasks, which may
result in potential challenges in integrating emotion and cause in real-world
applications. In this paper, we propose a Unified Multimodal Emotion
recognition and Emotion-Cause analysis framework (UniMEEC) to explore the
causality and complementarity between emotion and emotion cause. Concretely,
UniMEEC reformulates the MERC and MECPE tasks as two mask prediction problems,
enhancing the interaction between emotion and cause. Meanwhile, UniMEEC shares
the prompt learning among modalities for probing modality-specific knowledge
from the Pre-trained model. Furthermore, we propose a task-specific
hierarchical context aggregation to control the information flow to the task.
Experiment results on four public benchmark datasets verify the model
performance on MERC and MECPE tasks and achieve consistent improvements
compared with state-of-the-art methods. | cs.CL | UniMEEC: Towards Unified Multimodal Emotion Recognition and Emotion Cause |
2024-03-30T15:48:49Z | http://arxiv.org/pdf/2404.00401v1 | http://arxiv.org/abs/2404.00401v1 | Akash Ghosh, B Venkata Sahith, Niloy Ganguly, Pawan Goyal, Mayank Singh | 2024-03-30T15:48:49Z | cs.CL | Question-answering (QA) on hybrid scientific tabular and textual data deals
with scientific information, and relies on complex numerical reasoning. In
recent years, while tabular QA has seen rapid progress, understanding their
robustness on scientific information is lacking due to absence of any benchmark
dataset. To investigate the robustness of the existing state-of-the-art QA
models on scientific hybrid tabular data, we propose a new dataset, "SciTabQA",
consisting of 822 question-answer pairs from scientific tables and their
descriptions. With the help of this dataset, we assess the state-of-the-art
Tabular QA models based on their ability (i) to use heterogeneous information
requiring both structured data (table) and unstructured data (text) and (ii) to
perform complex scientific reasoning tasks. In essence, we check the capability
of the models to interpret scientific tables and text. Our experiments show
that "SciTabQA" is an innovative dataset to study question-answering over
scientific heterogeneous data. We benchmark three state-of-the-art Tabular QA
models, and find that the best F1 score is only 0.462. | cs.CL | How Robust are the Tabular QA Models for Scientific Tables? A Study using Customized Dataset |
2024-03-30T15:38:54Z | http://arxiv.org/pdf/2404.00399v2 | http://arxiv.org/abs/2404.00399v2 | Taishi Nakamura, Mayank Mishra, Simone Tedeschi, Yekun Chai, Jason T Stillerman, Felix Friedrich, Prateek Yadav, Tanmay Laud, Vu Minh Chien, Terry Yue Zhuo, Diganta Misra, Ben Bogin, Xuan-Son Vu, Marzena Karpinska, Arnav Varma Dantuluri, Wojciech Kusa, Tommaso Furlanello, Rio Yokota, Niklas Muennighoff, Suhas Pai, Tosin Adewumi, Veronika Laippala, Xiaozhe Yao, Adalberto Junior, Alpay Ariyak, Aleksandr Drozd, Jordan Clive, Kshitij Gupta, Liangyu Chen, Qi Sun, Ken Tsui, Noah Persaud, Nour Fahmy, Tianlong Chen, Mohit Bansal, Nicolo Monti, Tai Dang, Ziyang Luo, Tien-Tung Bui, Roberto Navigli, Virendra Mehta, Matthew Blumberg, Victor May, Huu Nguyen, Sampo Pyysalo | 2024-04-23T13:45:48Z | cs.CL, cs.AI, cs.LG | Pretrained language models underpin several AI applications, but their high
computational cost for training limits accessibility. Initiatives such as BLOOM
and StarCoder aim to democratize access to pretrained models for collaborative
community development. However, such existing models face challenges: limited
multilingual capabilities, continual pretraining causing catastrophic
forgetting, whereas pretraining from scratch is computationally expensive, and
compliance with AI safety and development laws. This paper presents Aurora-M, a
15B parameter multilingual open-source model trained on English, Finnish,
Hindi, Japanese, Vietnamese, and code. Continually pretrained from
StarCoderPlus on 435 billion additional tokens, Aurora-M surpasses 2 trillion
tokens in total training token count. It is the first open-source multilingual
model fine-tuned on human-reviewed safety instructions, thus aligning its
development not only with conventional red-teaming considerations, but also
with the specific concerns articulated in the Biden-Harris Executive Order on
the Safe, Secure, and Trustworthy Development and Use of Artificial
Intelligence. Aurora-M is rigorously evaluated across various tasks and
languages, demonstrating robustness against catastrophic forgetting and
outperforming alternatives in multilingual settings, particularly in safety
evaluations. To promote responsible open-source LLM development, Aurora-M and
its variants are released at
https://huggingface.co/collections/aurora-m/aurora-m-models-65fdfdff62471e09812f5407 . | cs.CL | Aurora-M: The First Open Source Multilingual Language Model Red-teamed according to the U.S. Executive Order |
2024-03-30T15:29:49Z | http://arxiv.org/pdf/2404.00397v1 | http://arxiv.org/abs/2404.00397v1 | Marco Cognetta, Tatsuya Hiraoka, Naoaki Okazaki, Rico Sennrich, Yuval Pinter | 2024-03-30T15:29:49Z | cs.CL | We explore threshold vocabulary trimming in Byte-Pair Encoding subword
tokenization, a postprocessing step that replaces rare subwords with their
component subwords. The technique is available in popular tokenization
libraries but has not been subjected to rigorous scientific scrutiny. While the
removal of rare subwords is suggested as best practice in machine translation
implementations, both as a means to reduce model size and for improving model
performance through robustness, our experiments indicate that, across a large
space of hyperparameter settings, vocabulary trimming fails to improve
performance, and is even prone to incurring heavy degradation. | cs.CL | An Analysis of BPE Vocabulary Trimming in Neural Machine Translation |
2024-03-30T14:58:44Z | http://arxiv.org/pdf/2404.00386v1 | http://arxiv.org/abs/2404.00386v1 | Parag Pravin Dakle, Alolika Gon, Sihan Zha, Liang Wang, SaiKrishna Rallabandi, Preethi Raghavan | 2024-03-30T14:58:44Z | cs.CL | In this paper, we describe the different approaches explored by the Jetsons
team for the Multi-Lingual ESG Impact Duration Inference (ML-ESG-3) shared
task. The shared task focuses on predicting the duration and type of the ESG
impact of a news article. The shared task dataset consists of 2,059 news titles
and articles in English, French, Korean, and Japanese languages. For the impact
duration classification task, we fine-tuned XLM-RoBERTa with a custom
fine-tuning strategy and using self-training and DeBERTa-v3 using only English
translations. These models individually ranked first on the leaderboard for
Korean and Japanese and in an ensemble for the English language, respectively.
For the impact type classification task, our XLM-RoBERTa model fine-tuned using
a custom fine-tuning strategy ranked first for the English language. | cs.CL | Jetsons at FinNLP 2024: Towards Understanding the ESG Impact of a News Article using Transformer-based Models |
2024-03-30T14:39:09Z | http://arxiv.org/pdf/2404.01336v2 | http://arxiv.org/abs/2404.01336v2 | Ziyi Zhou, Xiaoming Zhang, Litian Zhang, Jiacheng Liu, Xi Zhang, Chaozhuo Li | 2024-04-28T07:26:08Z | cs.CL, cs.AI, cs.MM | Existing benchmarks for fake news detection have significantly contributed to
the advancement of models in assessing the authenticity of news content.
However, these benchmarks typically focus solely on news pertaining to a single
semantic topic or originating from a single platform, thereby failing to
capture the diversity of multi-domain news in real scenarios. In order to
understand fake news across various domains, the external knowledge and
fine-grained annotations are indispensable to provide precise evidence and
uncover the diverse underlying strategies for fabrication, which are also
ignored by existing benchmarks. To address this gap, we introduce a novel
multi-domain knowledge-enhanced benchmark with fine-grained annotations, named
\textbf{FineFake}. FineFake encompasses 16,909 data samples spanning six
semantic topics and eight platforms. Each news item is enriched with
multi-modal content, potential social context, semi-manually verified common
knowledge, and fine-grained annotations that surpass conventional binary
labels. Furthermore, we formulate three challenging tasks based on FineFake and
propose a knowledge-enhanced domain adaptation network. Extensive experiments
are conducted on FineFake under various scenarios, providing accurate and
reliable benchmarks for future endeavors. The entire FineFake project is
publicly accessible as an open-source repository at
\url{https://github.com/Accuser907/FineFake}. | cs.CL | FineFake: A Knowledge-Enriched Dataset for Fine-Grained Multi-Domain Fake News Detection |
2024-03-30T14:09:00Z | http://arxiv.org/pdf/2404.00376v2 | http://arxiv.org/abs/2404.00376v2 | Hyunjae Kim, Hyeon Hwang, Jiwoo Lee, Sihyeon Park, Dain Kim, Taewhoo Lee, Chanwoong Yoon, Jiwoong Sohn, Donghee Choi, Jaewoo Kang | 2024-06-30T09:24:12Z | cs.CL | While recent advancements in commercial large language models (LM) have shown
promising results in medical tasks, their closed-source nature poses
significant privacy and security concerns, hindering their widespread use in
the medical field. Despite efforts to create open-source models, their limited
parameters often result in insufficient multi-step reasoning capabilities
required for solving complex medical problems. To address this, we introduce
Meerkat, a new family of medical AI systems ranging from 7 to 70 billion
parameters. The models were trained using our new synthetic dataset consisting
of high-quality chain-of-thought reasoning paths sourced from 18 medical
textbooks, along with diverse instruction-following datasets. Our systems
achieved remarkable accuracy across six medical benchmarks, surpassing the
previous best models such as MediTron and BioMistral, and GPT-3.5 by a large
margin. Notably, Meerkat-7B surpassed the passing threshold of the United
States Medical Licensing Examination (USMLE) for the first time for a
7B-parameter model, while Meerkat-70B outperformed GPT-4 by an average of 1.3%.
Additionally, Meerkat-70B correctly diagnosed 21 out of 38 complex clinical
cases, outperforming humans' 13.8 and closely matching GPT-4's 21.8. Our
systems offered more detailed free-form responses to clinical queries compared
to existing small models, approaching the performance level of large commercial
models. This significantly narrows the performance gap with large LMs,
showcasing its effectiveness in addressing complex medical challenges. | cs.CL | Small Language Models Learn Enhanced Reasoning Skills from Medical Textbooks |
2024-03-30T13:28:51Z | http://arxiv.org/pdf/2404.00361v1 | http://arxiv.org/abs/2404.00361v1 | Zhenhua Liu, Tong Zhu, Jianxiang Xiang, Wenliang Chen | 2024-03-30T13:28:51Z | cs.CL | Data augmentation (DA) is crucial to mitigate model training instability and
over-fitting problems in low-resource open-domain dialogue generation. However,
traditional DA methods often neglect semantic data diversity, restricting the
overall quality. Recently, large language models (LLM) have been used for DA to
generate diversified dialogues. However, they have limited controllability and
tend to generate dialogues with a distribution shift compared to the seed
dialogues. To maximize the augmentation diversity and address the
controllability problem, we propose \textbf{S}ummary-based \textbf{D}ialogue
\textbf{A}ugmentation with LLM (SDA). Our approach enhances the controllability
of LLM by using dialogue summaries as a planning tool. Based on summaries, SDA
can generate high-quality and diverse dialogue data even with a small seed
dataset. To evaluate the efficacy of data augmentation methods for open-domain
dialogue, we designed a clustering-based metric to characterize the semantic
diversity of the augmented dialogue data. The experimental results show that
SDA can augment high-quality and semantically diverse dialogues given a small
seed dataset and an LLM, and the augmented data can boost the performance of
open-domain dialogue models. | cs.CL | Controllable and Diverse Data Augmentation with Large Language Model for Low-Resource Open-Domain Dialogue Generation |
2024-03-30T12:48:31Z | http://arxiv.org/pdf/2404.00344v1 | http://arxiv.org/abs/2404.00344v1 | Ankit Satpute, Noah Giessing, Andre Greiner-Petter, Moritz Schubotz, Olaf Teschke, Akiko Aizawa, Bela Gipp | 2024-03-30T12:48:31Z | cs.CL, cs.AI, cs.IR | Large Language Models (LLMs) have demonstrated exceptional capabilities in
various natural language tasks, often achieving performances that surpass those
of humans. Despite these advancements, the domain of mathematics presents a
distinctive challenge, primarily due to its specialized structure and the
precision it demands. In this study, we adopted a two-step approach for
investigating the proficiency of LLMs in answering mathematical questions.
First, we employ the most effective LLMs, as identified by their performance on
math question-answer benchmarks, to generate answers to 78 questions from the
Math Stack Exchange (MSE). Second, a case analysis is conducted on the LLM that
showed the highest performance, focusing on the quality and accuracy of its
answers through manual evaluation. We found that GPT-4 performs best (nDCG of
0.48 and P@10 of 0.37) amongst existing LLMs fine-tuned for answering
mathematics questions and outperforms the current best approach on ArqMATH3
Task1, considering P@10. Our Case analysis indicates that while the GPT-4 can
generate relevant responses in certain instances, it does not consistently
answer all questions accurately. This paper explores the current limitations of
LLMs in navigating complex mathematical problem-solving. Through case analysis,
we shed light on the gaps in LLM capabilities within mathematics, thereby
setting the stage for future research and advancements in AI-driven
mathematical reasoning. We make our code and findings publicly available for
research: \url{https://github.com/gipplab/LLM-Investig-MathStackExchange} | cs.CL | Can LLMs Master Math? Investigating Large Language Models on Math Stack Exchange |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.