publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-02-19T07:48:25Z | http://arxiv.org/pdf/2402.11908v1 | http://arxiv.org/abs/2402.11908v1 | Sayeh Gholipour Picha, Dawood Al Chanti, Alice Caplier | 2024-02-19T07:48:25Z | cs.CL, cs.CV | Medical language processing and deep learning techniques have emerged as
critical tools for improving healthcare, particularly in the analysis of
medical imaging and medical text data. These multimodal data fusion techniques
help to improve the interpretation of medical imaging and lead to increased
diagnostic accuracy, informed clinical decisions, and improved patient
outcomes. The success of these models relies on the ability to extract and
consolidate semantic information from clinical text. This paper addresses the
need for more robust methods to evaluate the semantic content of medical
reports. Conventional natural language processing approaches and metrics are
initially designed for considering the semantic context in the natural language
domain and machine translation, often failing to capture the complex semantic
meanings inherent in medical content. In this study, we introduce a novel
approach designed specifically for assessing the semantic similarity between
generated medical reports and the ground truth. Our approach is validated,
demonstrating its efficiency in assessing domain-specific semantic similarity
within medical contexts. By applying our metric to state-of-the-art Chest X-ray
report generation models, we obtain results that not only align with
conventional metrics but also provide more contextually meaningful scores in
the considered medical domain. | cs.CL | Semantic Textual Similarity Assessment in Chest X-ray Reports Using a Domain-Specific Cosine-Based Metric |
2024-02-19T07:46:40Z | http://arxiv.org/pdf/2402.11907v1 | http://arxiv.org/abs/2402.11907v1 | Aiwei Liu, Haoping Bai, Zhiyun Lu, Xiang Kong, Simon Wang, Jiulong Shan, Meng Cao, Lijie Wen | 2024-02-19T07:46:40Z | cs.CL, 68T50, I.2.7 | Aligning large language models (LLMs) with human expectations without
human-annotated preference data is an important problem. In this paper, we
propose a method to evaluate the response preference by using the output
probabilities of response pairs under contrastive prompt pairs, which could
achieve better performance on LLaMA2-7B and LLaMA2-13B compared to RLAIF. Based
on this, we propose an automatic alignment method, Direct Large Model Alignment
(DLMA). First, we use contrastive prompt pairs to automatically generate
preference data. Then, we continue to evaluate the generated preference data
using contrastive prompt pairs and calculate a self-rewarding score. Finally,
we use the DPO algorithm to effectively align LLMs by combining this
self-rewarding score. In the experimental stage, our DLMA method could surpass
the \texttt{RLHF} method without relying on human-annotated preference data. | cs.CL | Direct Large Language Model Alignment Through Self-Rewarding Contrastive Prompt Distillation |
2024-02-19T07:45:17Z | http://arxiv.org/pdf/2402.11905v2 | http://arxiv.org/abs/2402.11905v2 | Yuxin Jiang, Yufei Wang, Chuhan Wu, Wanjun Zhong, Xingshan Zeng, Jiahui Gao, Liangyou Li, Xin Jiang, Lifeng Shang, Ruiming Tang, Qun Liu, Wei Wang | 2024-06-05T15:46:03Z | cs.CL | Knowledge editing techniques, aiming to efficiently modify a minor proportion
of knowledge in large language models (LLMs) without negatively impacting
performance across other inputs, have garnered widespread attention. However,
existing methods predominantly rely on memorizing the updated knowledge,
impeding LLMs from effectively combining the new knowledge with their inherent
knowledge when answering questions. To this end, we propose a Learning to Edit
(LTE) framework, focusing on teaching LLMs to apply updated knowledge into
input questions, inspired by the philosophy of "Teach a man to fish." LTE
features a two-phase process: (i) the Alignment Phase, which fine-tunes LLMs on
a meticulously curated parallel dataset to make reliable, in-scope edits while
preserving out-of-scope information and linguistic proficiency; and (ii) the
Inference Phase, which employs a retrieval-based mechanism for real-time and
mass knowledge editing. By comparing our approach with seven advanced baselines
across four popular knowledge editing benchmarks and two LLM architectures, we
demonstrate LTE's superiority in knowledge editing performance, robustness in
both batch and sequential editing, minimal interference on general tasks, and
rapid editing speeds. The data and code are available at
https://github.com/YJiangcm/LTE. | cs.CL | Learning to Edit: Aligning LLMs with Knowledge Editing |
2024-02-19T07:38:57Z | http://arxiv.org/pdf/2402.11903v3 | http://arxiv.org/abs/2402.11903v3 | Yu Zhang, Hui-Ling Zhen, Zehua Pei, Yingzhao Lian, Lihao Yin, Mingxuan Yuan, Bei Yu | 2024-06-19T02:52:00Z | cs.CL, cs.AI | Considering the challenges faced by large language models (LLMs) in logical
reasoning and planning, prior efforts have sought to augment LLMs with access
to external solvers. While progress has been made on simple reasoning problems,
solving classical constraint satisfaction problems, such as the Boolean
Satisfiability Problem (SAT) and Graph Coloring Problem (GCP), remains
difficult for off-the-shelf solvers due to their intricate expressions and
exponential search spaces. In this paper, we propose a novel differential logic
layer-aided language modeling (DiLA) approach, where logical constraints are
integrated into the forward and backward passes of a network layer, to provide
another option for LLM tool learning. In DiLA, LLM aims to transform the
language description to logic constraints and identify initial solutions of the
highest quality, while the differential logic layer focuses on iteratively
refining the LLM-prompted solution. Leveraging the logic layer as a bridge,
DiLA enhances the logical reasoning ability of LLMs on a range of reasoning
problems encoded by Boolean variables, guaranteeing the efficiency and
correctness of the solution process. We evaluate the performance of DiLA on two
classic reasoning problems and empirically demonstrate its consistent
outperformance against existing prompt-based and solver-aided approaches. | cs.CL | DiLA: Enhancing LLM Tool Learning with Differential Logic Layer |
2024-02-19T07:34:10Z | http://arxiv.org/pdf/2402.11900v2 | http://arxiv.org/abs/2402.11900v2 | Tianjie Ju, Yijin Chen, Xinwei Yuan, Zhuosheng Zhang, Wei Du, Yubin Zheng, Gongshen Liu | 2024-06-02T09:17:37Z | cs.CL | Recent work has showcased the powerful capability of large language models
(LLMs) in recalling knowledge and reasoning. However, the reliability of LLMs
in combining these two capabilities into reasoning through multi-hop facts has
not been widely explored. This paper systematically investigates the
possibilities for LLMs to utilize shortcuts based on direct connections between
the initial and terminal entities of multi-hop knowledge. We first explore the
existence of factual shortcuts through Knowledge Neurons, revealing that: (i)
the strength of factual shortcuts is highly correlated with the frequency of
co-occurrence of initial and terminal entities in the pre-training corpora;
(ii) few-shot prompting leverage more shortcuts in answering multi-hop
questions compared to chain-of-thought prompting. Then, we analyze the risks
posed by factual shortcuts from the perspective of multi-hop knowledge editing.
Analysis shows that approximately 20% of the failures are attributed to
shortcuts, and the initial and terminal entities in these failure instances
usually have higher co-occurrences in the pre-training corpus. Finally, we
propose erasing shortcut neurons to mitigate the associated risks and find that
this approach significantly reduces failures in multiple-hop knowledge editing
caused by shortcuts. | cs.CL | Investigating Multi-Hop Factual Shortcuts in Knowledge Editing of Large Language Models |
2024-02-19T07:22:29Z | http://arxiv.org/pdf/2402.11896v2 | http://arxiv.org/abs/2402.11896v2 | Zhihao Wen, Jie Zhang, Yuan Fang | 2024-06-01T07:13:15Z | cs.CL | Fine-tuning all parameters of large language models (LLMs) necessitates
substantial computational power and extended time. Latest advancements in
parameter-efficient fine-tuning (PEFT) techniques, such as Adapter tuning and
LoRA, allow for adjustments to only a minor fraction of the parameters of these
LLMs. Concurrently, it has been noted that the issue of over-smoothing
diminishes the effectiveness of these Transformer-based LLMs, resulting in
suboptimal performances in downstream tasks. In this paper, we present SIBO,
which is a SImple BOoster to enhance PEFT, by injecting an initial residual.
SIBO is straightforward and readily extensible to a range of state-of-the-art
PEFT techniques to alleviate over-smoothing and enhance performance. Extensive
experiments on 22 benchmark datasets demonstrate that SIBO significantly
enhances the performance of various strong baselines, achieving up to 15.7% and
23.5% improvement over existing PEFT methods on the arithmetic and commonsense
reasoning tasks, respectively. | cs.CL | SIBO: A Simple Booster for Parameter-Efficient Fine-Tuning |
2024-02-19T07:21:09Z | http://arxiv.org/pdf/2402.11895v3 | http://arxiv.org/abs/2402.11895v3 | Rochana Chaturvedi, Sugat Chaturvedi, Elena Zheleva | 2024-03-10T05:38:20Z | cs.SI, cs.CL, physics.soc-ph | While exposure to diverse viewpoints may reduce polarization, it can also
have a backfire effect and exacerbate polarization when the discussion is
adversarial. Here, we examine the question whether intergroup interactions
around important events affect polarization between majority and minority
groups in social networks. We compile data on the religious identity of nearly
700,000 Indian Twitter users engaging in COVID-19-related discourse during
2020. We introduce a new measure for an individual's group conformity based on
contextualized embeddings of tweet text, which helps us assess polarization
between religious groups. We then use a meta-learning framework to examine
heterogeneous treatment effects of intergroup interactions on an individual's
group conformity in the light of communal, political, and socio-economic
events. We find that for political and social events, intergroup interactions
reduce polarization. This decline is weaker for individuals at the extreme who
already exhibit high conformity to their group. In contrast, during communal
events, intergroup interactions can increase group conformity. Finally, we
decompose the differential effects across religious groups in terms of emotions
and topics of discussion. The results show that the dynamics of religious
polarization are sensitive to the context and have important implications for
understanding the role of intergroup interactions. | cs.SI | Bridging or Breaking: Impact of Intergroup Interactions on Religious Polarization |
2024-02-19T07:15:59Z | http://arxiv.org/pdf/2402.11894v3 | http://arxiv.org/abs/2402.11894v3 | Jiahao Ying, Yixin Cao, Yushi Bai, Qianru Sun, Bo Wang, Wei Tang, Zhaojun Ding, Yizhe Yang, Xuanjing Huang, Shuicheng Yan | 2024-06-06T12:49:44Z | cs.CL | Large language models (LLMs) have achieved impressive performance across
various natural language benchmarks, prompting a continual need to curate more
difficult datasets for larger LLMs, which is costly and time-consuming. In this
paper, we propose to automate dataset updating and provide systematic analysis
regarding its effectiveness in dealing with benchmark leakage issue, difficulty
control, and stability. Thus, once the current benchmark has been mastered or
leaked, we can update it for timely and reliable evaluation. There are two
updating strategies: 1) mimicking strategy to generate similar samples based on
original data, preserving stylistic and contextual essence, and 2) extending
strategy that further expands existing samples at varying cognitive levels by
adapting Bloom's taxonomy of educational objectives. Extensive experiments on
updated MMLU and BIG-Bench demonstrate the stability of the proposed strategies
and find that the mimicking strategy can effectively alleviate issues of
overestimation from benchmark leakage. In cases where the efficient mimicking
strategy fails, our extending strategy still shows promising results.
Additionally, by controlling the difficulty, we can better discern the models'
performance and enable fine-grained analysis neither too difficult nor too easy
an exam can fairly judge students' learning status. To the best of our
knowledge, we are the first to automate updating benchmarks for reliable and
timely evaluation. Our demo leaderboard can be found at
https://yingjiahao14.github.io/Automating-DatasetUpdates/. | cs.CL | Automating Dataset Updates Towards Reliable and Timely Evaluation of Large Language Models |
2024-02-19T07:06:52Z | http://arxiv.org/pdf/2402.11891v1 | http://arxiv.org/abs/2402.11891v1 | Shuai Wang, Ekaterina Khramtsova, Shengyao Zhuang, Guido Zuccon | 2024-02-19T07:06:52Z | cs.IR, cs.CL | Federated search systems aggregate results from multiple search engines,
selecting appropriate sources to enhance result quality and align with user
intent. With the increasing uptake of Retrieval-Augmented Generation (RAG)
pipelines, federated search can play a pivotal role in sourcing relevant
information across heterogeneous data sources to generate informed responses.
However, existing datasets, such as those developed in the past TREC FedWeb
tracks, predate the RAG paradigm shift and lack representation of modern
information retrieval challenges. To bridge this gap, we present FeB4RAG, a
novel dataset specifically designed for federated search within RAG frameworks.
This dataset, derived from 16 sub-collections of the widely used \beir
benchmarking collection, includes 790 information requests (akin to
conversational queries) tailored for chatbot applications, along with top
results returned by each resource and associated LLM-derived relevance
judgements. Additionally, to support the need for this collection, we
demonstrate the impact on response generation of a high quality federated
search system for RAG compared to a naive approach to federated search. We do
so by comparing answers generated through the RAG pipeline through a
qualitative side-by-side comparison. Our collection fosters and supports the
development and evaluation of new federated search methods, especially in the
context of RAG pipelines. | cs.IR | FeB4RAG: Evaluating Federated Search in the Context of Retrieval Augmented Generation |
2024-02-19T07:01:10Z | http://arxiv.org/pdf/2402.11890v2 | http://arxiv.org/abs/2402.11890v2 | Qihuang Zhong, Liang Ding, Li Shen, Juhua Liu, Bo Du, Dacheng Tao | 2024-06-17T02:44:28Z | cs.CL | Knowledge distillation (KD) is a common approach to compress a teacher model
to reduce its inference cost and memory footprint, by training a smaller
student model. However, in the context of autoregressive language models (LMs),
we empirically find that larger teacher LMs might dramatically result in a
poorer student. In response to this problem, we conduct a series of analyses
and reveal that different tokens have different teaching modes, neglecting
which will lead to performance degradation. Motivated by this, we propose a
simple yet effective adaptive teaching approach (ATKD) to improve the KD. The
core of ATKD is to reduce rote learning and make teaching more diverse and
flexible. Extensive experiments on 8 LM tasks show that, with the help of ATKD,
various baseline KD methods can achieve consistent and significant performance
gains (up to +3.04% average score) across all model types and sizes. More
encouragingly, ATKD can improve the student model generalization effectively. | cs.CL | Revisiting Knowledge Distillation for Autoregressive Language Models |
2024-02-19T06:58:42Z | http://arxiv.org/pdf/2402.11889v2 | http://arxiv.org/abs/2402.11889v2 | Qihuang Zhong, Liang Ding, Juhua Liu, Bo Du, Dacheng Tao | 2024-06-17T02:48:21Z | cs.CL | With the development of instruction-tuned large language models (LLMs),
improving the safety of LLMs has become more critical. However, the current
approaches for aligning the LLMs output with expected safety usually require
substantial training efforts, e.g., high-quality safety data and expensive
computational resources, which are costly and inefficient. To this end, we
present reverse prompt contrastive decoding (ROSE), a simple-yet-effective
method to directly boost the safety of existing instruction-tuned LLMs without
any additional training. The principle of ROSE is to improve the probability of
desired safe output via suppressing the undesired output induced by the
carefully-designed reverse prompts. Experiments on 6 safety and 2
general-purpose tasks show that, our ROSE not only brings consistent and
significant safety improvements (up to +13.8% safety score) upon 5 types of
instruction-tuned LLMs, but also benefits the general-purpose ability of LLMs.
In-depth analyses explore the underlying mechanism of ROSE, and reveal when and
where to use it. | cs.CL | ROSE Doesn't Do That: Boosting the Safety of Instruction-Tuned Large Language Models with Reverse Prompt Contrastive Decoding |
2024-02-19T06:57:02Z | http://arxiv.org/pdf/2402.14840v1 | http://arxiv.org/abs/2402.14840v1 | Congyun Jin, Ming Zhang, Xiaowei Ma, Li Yujiao, Yingbo Wang, Yabo Jia, Yuliang Du, Tao Sun, Haowen Wang, Cong Fan, Jinjie Gu, Chenfei Chi, Xiangguo Lv, Fangzhou Li, Wei Xue, Yiran Huang | 2024-02-19T06:57:02Z | cs.CL, cs.AI, stat.AP | Recent advancements in Large Language Models (LLMs) and Large Multi-modal
Models (LMMs) have shown potential in various medical applications, such as
Intelligent Medical Diagnosis. Although impressive results have been achieved,
we find that existing benchmarks do not reflect the complexity of real medical
reports and specialized in-depth reasoning capabilities. In this work, we
introduced RJUA-MedDQA, a comprehensive benchmark in the field of medical
specialization, which poses several challenges: comprehensively interpreting
imgage content across diverse challenging layouts, possessing numerical
reasoning ability to identify abnormal indicators and demonstrating clinical
reasoning ability to provide statements of disease diagnosis, status and advice
based on medical contexts. We carefully design the data generation pipeline and
proposed the Efficient Structural Restoration Annotation (ESRA) Method, aimed
at restoring textual and tabular content in medical report images. This method
substantially enhances annotation efficiency, doubling the productivity of each
annotator, and yields a 26.8% improvement in accuracy. We conduct extensive
evaluations, including few-shot assessments of 5 LMMs which are capable of
solving Chinese medical QA tasks. To further investigate the limitations and
potential of current LMMs, we conduct comparative experiments on a set of
strong LLMs by using image-text generated by ESRA method. We report the
performance of baselines and offer several observations: (1) The overall
performance of existing LMMs is still limited; however LMMs more robust to
low-quality and diverse-structured images compared to LLMs. (3) Reasoning
across context and image content present significant challenges. We hope this
benchmark helps the community make progress on these challenging tasks in
multi-modal medical document understanding and facilitate its application in
healthcare. | cs.CL | RJUA-MedDQA: A Multimodal Benchmark for Medical Document Question Answering and Clinical Reasoning |
2024-02-19T06:54:55Z | http://arxiv.org/pdf/2402.11886v1 | http://arxiv.org/abs/2402.11886v1 | Shir Lissak, Nitay Calderon, Geva Shenkman, Yaakov Ophir, Eyal Fruchter, Anat Brunstein Klomek, Roi Reichart | 2024-02-19T06:54:55Z | cs.CL, cs.AI | Queer youth face increased mental health risks, such as depression, anxiety,
and suicidal ideation. Hindered by negative stigma, they often avoid seeking
help and rely on online resources, which may provide incompatible information.
Although access to a supportive environment and reliable information is
invaluable, many queer youth worldwide have no access to such support. However,
this could soon change due to the rapid adoption of Large Language Models
(LLMs) such as ChatGPT. This paper aims to comprehensively explore the
potential of LLMs to revolutionize emotional support for queers. To this end,
we conduct a qualitative and quantitative analysis of LLM's interactions with
queer-related content. To evaluate response quality, we develop a novel
ten-question scale that is inspired by psychological standards and expert
input. We apply this scale to score several LLMs and human comments to posts
where queer youth seek advice and share experiences. We find that LLM responses
are supportive and inclusive, outscoring humans. However, they tend to be
generic, not empathetic enough, and lack personalization, resulting in
nonreliable and potentially harmful advice. We discuss these challenges,
demonstrate that a dedicated prompt can improve the performance, and propose a
blueprint of an LLM-supporter that actively (but sensitively) seeks user
context to provide personalized, empathetic, and reliable responses. Our
annotated dataset is available for further research. | cs.CL | The Colorful Future of LLMs: Evaluating and Improving LLMs as Emotional Supporters for Queer Youth |
2024-02-19T06:32:39Z | http://arxiv.org/pdf/2402.11875v1 | http://arxiv.org/abs/2402.11875v1 | Hongcheng Liu, Pingjie Wang, Yu Wang, Yanfeng Wang | 2024-02-19T06:32:39Z | cs.CL | Video-grounded dialogue generation (VDG) requires the system to generate a
fluent and accurate answer based on multimodal knowledge. However, the
difficulty in multimodal knowledge utilization brings serious hallucinations to
VDG models in practice. Although previous works mitigate the hallucination in a
variety of ways, they hardly take notice of the importance of the multimodal
knowledge anchor answer tokens. In this paper, we reveal via perplexity that
different VDG models experience varying hallucinations and exhibit diverse
anchor tokens. Based on this observation, we propose M2K-VDG, a model-adaptive
multimodal knowledge anchor enhancement framework for hallucination reduction.
Furthermore, we introduce the counterfactual effect for more accurate anchor
token detection. The experimental results on three popular benchmarks exhibit
the superiority of our approach over state-of-the-art methods, demonstrating
its effectiveness in reducing hallucinations. | cs.CL | M2K-VDG: Model-Adaptive Multimodal Knowledge Anchor Enhanced Video-grounded Dialogue Generation |
2024-02-19T06:11:28Z | http://arxiv.org/pdf/2402.11863v3 | http://arxiv.org/abs/2402.11863v3 | Wei Jie Yeo, Ranjan Satapathy, Rick Siow Mong Goh, Erik Cambria | 2024-04-01T08:33:11Z | cs.CL | Prompt Engineering has garnered significant attention for enhancing the
performance of large language models across a multitude of tasks. Techniques
such as the Chain-of-Thought not only bolster task performance but also
delineate a clear trajectory of reasoning steps, offering a tangible form of
explanation for the audience. Prior works on interpretability assess the
reasoning chains yielded by Chain-of-Thought solely along a singular axis,
namely faithfulness. We present a comprehensive and multifaceted evaluation of
interpretability, examining not only faithfulness but also robustness and
utility across multiple commonsense reasoning benchmarks. Likewise, our
investigation is not confined to a single prompting technique; it expansively
covers a multitude of prevalent prompting techniques employed in large language
models, thereby ensuring a wide-ranging and exhaustive evaluation. In addition,
we introduce a simple interpretability alignment technique, termed
Self-Entailment-Alignment Chain-of-thought, that yields more than 70\%
improvements across multiple dimensions of interpretability. Code is available
at https://github.com/SenticNet/CoT_interpretability | cs.CL | How Interpretable are Reasoning Explanations from Prompting Large Language Models? |
2024-02-19T05:15:13Z | http://arxiv.org/pdf/2402.11845v1 | http://arxiv.org/abs/2402.11845v1 | Rui Cao, Roy Ka-Wei Lee, Jing Jiang | 2024-02-19T05:15:13Z | cs.CL, cs.CV | In this paper, we address the challenge of detecting hateful memes in the
low-resource setting where only a few labeled examples are available. Our
approach leverages the compositionality of Low-rank adaptation (LoRA), a widely
used parameter-efficient tuning technique. We commence by fine-tuning large
language models (LLMs) with LoRA on selected tasks pertinent to hateful meme
detection, thereby generating a suite of LoRA modules. These modules are
capable of essential reasoning skills for hateful meme detection. We then use
the few available annotated samples to train a module composer, which assigns
weights to the LoRA modules based on their relevance. The model's learnable
parameters are directly proportional to the number of LoRA modules. This
modularized network, underpinned by LLMs and augmented with LoRA modules,
exhibits enhanced generalization in the context of hateful meme detection. Our
evaluation spans three datasets designed for hateful meme detection in a
few-shot learning context. The proposed method demonstrates superior
performance to traditional in-context learning, which is also more
computationally intensive during inference.We then use the few available
annotated samples to train a module composer, which assigns weights to the LoRA
modules based on their relevance. The model's learnable parameters are directly
proportional to the number of LoRA modules. This modularized network,
underpinned by LLMs and augmented with LoRA modules, exhibits enhanced
generalization in the context of hateful meme detection. Our evaluation spans
three datasets designed for hateful meme detection in a few-shot learning
context. The proposed method demonstrates superior performance to traditional
in-context learning, which is also more computationally intensive during
inference. | cs.CL | Modularized Networks for Few-shot Hateful Meme Detection |
2024-02-19T05:06:10Z | http://arxiv.org/pdf/2402.11839v1 | http://arxiv.org/abs/2402.11839v1 | Mahsa Azarshab, Mohammad Fathian, Babak Amiri | 2024-02-19T05:06:10Z | cs.NE, cs.CL, cs.LG | Text document clustering can play a vital role in organizing and handling the
everincreasing number of text documents. Uninformative and redundant features
included in large text documents reduce the effectiveness of the clustering
algorithm. Feature selection (FS) is a well-known technique for removing these
features. Since FS can be formulated as an optimization problem, various
meta-heuristic algorithms have been employed to solve it.
Teaching-Learning-Based Optimization (TLBO) is a novel meta-heuristic algorithm
that benefits from the low number of parameters and fast convergence. A hybrid
method can simultaneously benefit from the advantages of TLBO and tackle the
possible entrapment in the local optimum. By proposing a hybrid of TLBO, Grey
Wolf Optimizer (GWO), and Genetic Algorithm (GA) operators, this paper suggests
a filter-based FS algorithm (TLBO-GWO). Six benchmark datasets are selected,
and TLBO-GWO is compared with three recently proposed FS algorithms with
similar approaches, the main TLBO and GWO. The comparison is conducted based on
clustering evaluation measures, convergence behavior, and dimension reduction,
and is validated using statistical tests. The results reveal that TLBO-GWO can
significantly enhance the effectiveness of the text clustering technique
(K-means). | cs.NE | An enhanced Teaching-Learning-Based Optimization (TLBO) with Grey Wolf Optimizer (GWO) for text feature selection and clustering |
2024-02-19T04:41:31Z | http://arxiv.org/pdf/2402.11827v1 | http://arxiv.org/abs/2402.11827v1 | Chanwoong Yoon, Gangwoo Kim, Byeongguk Jeon, Sungdong Kim, Yohan Jo, Jaewoo Kang | 2024-02-19T04:41:31Z | cs.IR, cs.CL | Conversational search, unlike single-turn retrieval tasks, requires
understanding the current question within a dialogue context. The common
approach of rewrite-then-retrieve aims to decontextualize questions to be
self-sufficient for off-the-shelf retrievers, but most existing methods produce
sub-optimal query rewrites due to the limited ability to incorporate signals
from the retrieval results. To overcome this limitation, we present a novel
framework RetPO (Retriever's Preference Optimization), which is designed to
optimize a language model (LM) for reformulating search queries in line with
the preferences of the target retrieval systems. The process begins by
prompting a large LM to produce various potential rewrites and then collects
retrieval performance for these rewrites as the retrievers' preferences.
Through the process, we construct a large-scale dataset called RF collection,
containing Retrievers' Feedback on over 410K query rewrites across 12K
conversations. Furthermore, we fine-tune a smaller LM using this dataset to
align it with the retrievers' preferences as feedback. The resulting model
achieves state-of-the-art performance on two recent conversational search
benchmarks, significantly outperforming existing baselines, including GPT-3.5. | cs.IR | Ask Optimal Questions: Aligning Large Language Models with Retriever's Preference in Conversational Search |
2024-02-19T04:29:45Z | http://arxiv.org/pdf/2402.11821v2 | http://arxiv.org/abs/2402.11821v2 | Yanbang Wang, Hejie Cui, Jon Kleinberg | 2024-02-20T03:23:49Z | cs.LG, cs.CL, cs.IR, cs.SI | Graphs data is crucial for many applications, and much of it exists in the
relations described in textual format. As a result, being able to accurately
recall and encode a graph described in earlier text is a basic yet pivotal
ability that LLMs need to demonstrate if they are to perform reasoning tasks
that involve graph-structured information. Human performance at graph recall
has been studied by cognitive scientists for decades, and has been found to
often exhibit certain structural patterns of bias that align with human
handling of social relationships. To date, however, we know little about how
LLMs behave in analogous graph recall tasks: do their recalled graphs also
exhibit certain biased patterns, and if so, how do they compare with humans and
affect other graph reasoning tasks? In this work, we perform the first
systematical study of graph recall by LLMs, investigating the accuracy and
biased microstructures (local structural patterns) in their recall. We find
that LLMs not only underperform often in graph recall, but also tend to favor
more triangles and alternating 2-paths. Moreover, we find that more advanced
LLMs have a striking dependence on the domain that a real-world graph comes
from -- by yielding the best recall accuracy when the graph is narrated in a
language style consistent with its original domain. | cs.LG | Microstructures and Accuracy of Graph Recall by Large Language Models |
2024-02-19T04:19:36Z | http://arxiv.org/pdf/2402.11819v1 | http://arxiv.org/abs/2402.11819v1 | Zouying Cao, Yifei Yang, Hai Zhao | 2024-02-19T04:19:36Z | cs.CL | Large Language Models (LLMs) suffer from huge number of parameters, which
restricts their deployment on edge devices. Weight sharing is one promising
solution that encourages weight reuse, effectively reducing memory usage with
less performance drop. However, current weight sharing techniques primarily
focus on small-scale models like BERT and employ coarse-grained sharing rules,
e.g., layer-wise. This becomes limiting given the prevalence of LLMs and
sharing an entire layer or block obviously diminishes the flexibility of weight
sharing. In this paper, we present a perspective on $\textit{$\textbf{head-wise
shareable attention for large language models}$}$. We further propose two
memory-efficient methods that share parameters across attention heads, with a
specific focus on LLMs. Both of them use the same dynamic strategy to select
the shared weight matrices. The first method directly reuses the pre-trained
weights without retraining, denoted as $\textbf{DirectShare}$. The second
method first post-trains with constraint on weight matrix similarity and then
shares, denoted as $\textbf{PostShare}$. Experimental results reveal our
head-wise shared models still maintain satisfactory capabilities, demonstrating
the feasibility of fine-grained weight sharing applied to LLMs. | cs.CL | Head-wise Shareable Attention for Large Language Models |
2024-02-19T04:17:21Z | http://arxiv.org/pdf/2402.11818v1 | http://arxiv.org/abs/2402.11818v1 | Sameer Jain, Sedrick Scott Keh, Shova Chettri, Karun Dewan, Pablo Izquierdo, Johanna Prussman, Pooja Shreshtha, Cesar Suarez, Zheyuan Ryan Shi, Lei Li, Fei Fang | 2024-02-19T04:17:21Z | cs.CL, cs.AI, cs.CY | Environmental conservation organizations routinely monitor news content on
conservation in protected areas to maintain situational awareness of
developments that can have an environmental impact. Existing automated media
monitoring systems require large amounts of data labeled by domain experts,
which is only feasible at scale for high-resource languages like English.
However, such tools are most needed in the global south where news of interest
is mainly in local low-resource languages, and far fewer experts are available
to annotate datasets sustainably. In this paper, we propose NewsSerow, a method
to automatically recognize environmental conservation content in low-resource
languages. NewsSerow is a pipeline of summarization, in-context few-shot
classification, and self-reflection using large language models (LLMs). Using
at most 10 demonstration example news articles in Nepali, NewsSerow
significantly outperforms other few-shot methods and achieves comparable
performance with models fully fine-tuned using thousands of examples. The World
Wide Fund for Nature (WWF) has deployed NewsSerow for media monitoring in
Nepal, significantly reducing their operational burden, and ensuring that AI
tools for conservation actually reach the communities that need them the most.
NewsSerow has also been deployed for countries with other languages like
Colombia. | cs.CL | Where It Really Matters: Few-Shot Environmental Conservation Media Monitoring for Low-Resource Languages |
2024-02-19T04:11:34Z | http://arxiv.org/pdf/2402.11815v2 | http://arxiv.org/abs/2402.11815v2 | Shubhashis Roy Dipta, Sadat Shahriar | 2024-03-27T20:30:08Z | cs.CL, cs.AI, cs.LG | This paper describes our system developed for SemEval-2024 Task 8,
``Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated
Text Detection'' Machine-generated texts have been one of the main concerns due
to the use of large language models (LLM) in fake text generation, phishing,
cheating in exams, or even plagiarizing copyright materials. A lot of systems
have been developed to detect machine-generated text. Nonetheless, the majority
of these systems rely on the text-generating model. This limitation is
impractical in real-world scenarios, as it's often impossible to know which
specific model the user has used for text generation. In this work, we propose
a $\textbf{single}$ model based on contrastive learning, which uses
$\textbf{$\approx$40% of the baseline's parameters}$ (149M vs. 355M) but shows
a comparable performance on the test dataset $(\textbf{21st out of 137
participants})$. Our key finding is that even without an ensemble of multiple
models, a single base model can have comparable performance with the help of
data augmentation and contrastive learning. Our code is publicly available at
https://github.com/dipta007/SemEval24-Task8. | cs.CL | HU at SemEval-2024 Task 8A: Can Contrastive Learning Learn Embeddings to Detect Machine-Generated Text? |
2024-02-19T03:56:44Z | http://arxiv.org/pdf/2402.11811v2 | http://arxiv.org/abs/2402.11811v2 | Junru Lu, Siyu An, Min Zhang, Yulan He, Di Yin, Xing Sun | 2024-06-16T10:29:03Z | cs.CL | When the quality of naive prompts is carefully optimized by human experts,
the task performance of large language models (LLMs) can be significantly
improved. However, expert-based prompt optimizations are expensive. Herein,
some works have proposed Automatic Prompt Optimization (APO), to optimize naive
prompts according to task outputs of given in-box testing models, with the help
of advanced LLMs (e.g., GPT-4) in an ad-hoc way. Although effective, existing
schemes suffer from poor generalization ability and privacy risk. To this end,
we collect the first large-scale Prompt Optimization Preference dataset (POP),
fine-tune offline local LLM-based optimizers, then fairly test with various
downstream models. Our method allows accurate optimization of the core task
instruction part within the naive prompt in a model-agnostic manner, and thus
is named Free-from Instruction-oriented Prompt Optimization (FIPO). In
specific, FIPO uses a modular APO template that dynamically integrate the naive
task instruction, optional instruction responses, and optional ground truth to
produce finely optimized prompts. The POP dataset is meticulously constructed
using advanced LLMs, undergoing rigorous cross-validation by human experts and
analytical models. Leveraging insights from the data with Tulu2 models and
diverse fine-tuning strategies, we validate the efficacy of FIPO framework
across five public benchmarks and three testing models. Check codes and data
here: https://github.com/LuJunru/FIPO_Project. | cs.CL | FIPO: Free-form Instruction-oriented Prompt Optimization with Preference Dataset and Modular Fine-tuning Schema |
2024-02-19T03:39:10Z | http://arxiv.org/pdf/2402.11809v3 | http://arxiv.org/abs/2402.11809v3 | Hanling Yi, Feng Lin, Hongbin Li, Peiyang Ning, Xiaotian Yu, Rong Xiao | 2024-05-20T01:48:18Z | cs.CL, cs.AI, cs.LG | This research aims to accelerate the inference speed of large language models
(LLMs) with billions of parameters. We propose \textbf{S}mart \textbf{P}arallel
\textbf{A}uto-\textbf{C}orrect d\textbf{E}coding (SPACE), an innovative
approach designed for achieving lossless acceleration of LLMs. By integrating
semi-autoregressive inference and speculative decoding capabilities, SPACE
uniquely enables autoregressive LLMs to parallelize token generation and
verification. This is realized through a specialized semi-autoregressive
supervised fine-tuning process that equips existing LLMs with the ability to
simultaneously predict multiple tokens. Additionally, an auto-correct decoding
algorithm facilitates the simultaneous generation and verification of token
sequences within a single model invocation. Through extensive experiments on a
range of LLMs, SPACE has demonstrated inference speedup ranging from 2.7x-4.0x
on HumanEval-X while maintaining output quality. | cs.CL | Generation Meets Verification: Accelerating Large Language Model Inference with Smart Parallel Auto-Correct Decoding |
2024-02-19T03:21:19Z | http://arxiv.org/pdf/2402.11804v3 | http://arxiv.org/abs/2402.11804v3 | Kai Wang, Yuwei Xu, Zhiyong Wu, Siqiang Luo | 2024-06-19T09:00:53Z | cs.AI, cs.CL, cs.SI | Knowledge Graph (KG) inductive reasoning, which aims to infer missing facts
from new KGs that are not seen during training, has been widely adopted in
various applications. One critical challenge of KG inductive reasoning is
handling low-resource scenarios with scarcity in both textual and structural
aspects. In this paper, we attempt to address this challenge with Large
Language Models (LLMs). Particularly, we utilize the state-of-the-art LLMs to
generate a graph-structural prompt to enhance the pre-trained Graph Neural
Networks (GNNs), which brings us new methodological insights into the KG
inductive reasoning methods, as well as high generalizability in practice. On
the methodological side, we introduce a novel pretraining and prompting
framework ProLINK, designed for low-resource inductive reasoning across
arbitrary KGs without requiring additional training. On the practical side, we
experimentally evaluate our approach on 36 low-resource KG datasets and find
that ProLINK outperforms previous methods in three-shot, one-shot, and
zero-shot reasoning tasks, exhibiting average performance improvements by 20%,
45%, and 147%, respectively. Furthermore, ProLINK demonstrates strong
robustness for various LLM promptings as well as full-shot scenarios. | cs.AI | LLM as Prompter: Low-resource Inductive Reasoning on Arbitrary Knowledge Graphs |
2024-02-19T02:48:44Z | http://arxiv.org/pdf/2402.11794v1 | http://arxiv.org/abs/2402.11794v1 | Zizhong Li, Haopeng Zhang, Jiawei Zhang | 2024-02-19T02:48:44Z | cs.CL, cs.IR | Retrieval-augmented generation framework can address the limitations of large
language models by enabling real-time knowledge updates for more accurate
answers. An efficient way in the training phase of retrieval-augmented models
is attention distillation, which uses attention scores as a supervision signal
instead of manually annotated query-document pairs. Despite its growing
popularity, the detailed mechanisms behind the success of attention
distillation remain unexplored, particularly the specific patterns it leverages
to benefit training. In this paper, we address this gap by conducting a
comprehensive review of attention distillation workflow and identifying key
factors influencing the learning quality of retrieval-augmented language
models. We further propose indicators for optimizing models' training methods
and avoiding ineffective training. | cs.CL | Unveiling the Magic: Investigating Attention Distillation in Retrieval-augmented Generation |
2024-02-19T02:15:34Z | http://arxiv.org/pdf/2402.11782v1 | http://arxiv.org/abs/2402.11782v1 | Alexander Wan, Eric Wallace, Dan Klein | 2024-02-19T02:15:34Z | cs.CL, cs.LG | Retrieval-augmented language models are being increasingly tasked with
subjective, contentious, and conflicting queries such as "is aspartame linked
to cancer". To resolve these ambiguous queries, one must search through a large
range of websites and consider "which, if any, of this evidence do I find
convincing?". In this work, we study how LLMs answer this question. In
particular, we construct ConflictingQA, a dataset that pairs controversial
queries with a series of real-world evidence documents that contain different
facts (e.g., quantitative results), argument styles (e.g., appeals to
authority), and answers (Yes or No). We use this dataset to perform sensitivity
and counterfactual analyses to explore which text features most affect LLM
predictions. Overall, we find that current models rely heavily on the relevance
of a website to the query, while largely ignoring stylistic features that
humans find important such as whether a text contains scientific references or
is written with a neutral tone. Taken together, these results highlight the
importance of RAG corpus quality (e.g., the need to filter misinformation), and
possibly even a shift in how LLMs are trained to better align with human
judgements. | cs.CL | What Evidence Do Language Models Find Convincing? |
2024-02-19T02:08:03Z | http://arxiv.org/pdf/2402.11777v1 | http://arxiv.org/abs/2402.11777v1 | Pedro Freire, ChengCheng Tan, Adam Gleave, Dan Hendrycks, Scott Emmons | 2024-02-19T02:08:03Z | cs.CL, cs.AI, cs.LG, I.2.7 | Do language models implicitly learn a concept of human wellbeing? We explore
this through the ETHICS Utilitarianism task, assessing if scaling enhances
pretrained models' representations. Our initial finding reveals that, without
any prompt engineering or finetuning, the leading principal component from
OpenAI's text-embedding-ada-002 achieves 73.9% accuracy. This closely matches
the 74.6% of BERT-large finetuned on the entire ETHICS dataset, suggesting
pretraining conveys some understanding about human wellbeing. Next, we consider
four language model families, observing how Utilitarianism accuracy varies with
increased parameters. We find performance is nondecreasing with increased model
size when using sufficient numbers of principal components. | cs.CL | Uncovering Latent Human Wellbeing in Language Model Embeddings |
2024-02-19T01:49:53Z | http://arxiv.org/pdf/2402.11770v2 | http://arxiv.org/abs/2402.11770v2 | Md Arafat Sultan, Jatin Ganhotra, Ramón Fernandez Astudillo | 2024-02-20T02:55:57Z | cs.CL | We introduce a structured chain-of-thought (SCoT) prompting approach to
generating content-grounded multi-turn question-answer conversations using a
pre-trained large language model (LLM). At the core of our proposal is a
structured breakdown of the complex task into a number of states in a state
machine, so that actions corresponding to various subtasks, e.g., content
reading and utterance generation, can be executed in their own dedicated
states. Each state leverages a unique set of resources including prompts and
(optionally) additional tools to augment the generation process. Our
experimental results show that SCoT prompting with designated states for
hallucination mitigation increases agent faithfulness to grounding documents by
up to 16.8%. When used as training data, our open-domain conversations
synthesized from only 6 Wikipedia-based seed demonstrations train strong
conversational QA agents; in out-of-domain evaluation, for example, we observe
improvements of up to 13.9% over target domain gold data when the latter is
augmented with our generated examples. | cs.CL | Structured Chain-of-Thought Prompting for Few-Shot Generation of Content-Grounded QA Conversations |
2024-02-19T01:28:48Z | http://arxiv.org/pdf/2402.11764v1 | http://arxiv.org/abs/2402.11764v1 | Pengrui Han, Rafal Kocielnik, Adhithya Saravanan, Roy Jiang, Or Sharir, Anima Anandkumar | 2024-02-19T01:28:48Z | cs.CL, cs.AI, cs.CY, 68T50, I.2.7; K.4.1 | Large Language models (LLMs), while powerful, exhibit harmful social biases.
Debiasing is often challenging due to computational costs, data constraints,
and potential degradation of multi-task language capabilities. This work
introduces a novel approach utilizing ChatGPT to generate synthetic training
data, aiming to enhance the debiasing of LLMs. We propose two strategies:
Targeted Prompting, which provides effective debiasing for known biases but
necessitates prior specification of bias in question; and General Prompting,
which, while slightly less effective, offers debiasing across various
categories. We leverage resource-efficient LLM debiasing using adapter tuning
and compare the effectiveness of our synthetic data to existing debiasing
datasets. Our results reveal that: (1) ChatGPT can efficiently produce
high-quality training data for debiasing other LLMs; (2) data produced via our
approach surpasses existing datasets in debiasing performance while also
preserving internal knowledge of a pre-trained LLM; and (3) synthetic data
exhibits generalizability across categories, effectively mitigating various
biases, including intersectional ones. These findings underscore the potential
of synthetic data in advancing the fairness of LLMs with minimal retraining
cost. | cs.CL | ChatGPT Based Data Augmentation for Improved Parameter-Efficient Debiasing of LLMs |
2024-02-19T01:11:44Z | http://arxiv.org/pdf/2402.11757v1 | http://arxiv.org/abs/2402.11757v1 | Shuai Wang, Shengyao Zhuang, Guido Zuccon | 2024-02-19T01:11:44Z | cs.IR, cs.CL | Text stemming is a natural language processing technique that is used to
reduce words to their base form, also known as the root form. The use of
stemming in IR has been shown to often improve the effectiveness of
keyword-matching models such as BM25. However, traditional stemming methods,
focusing solely on individual terms, overlook the richness of contextual
information. Recognizing this gap, in this paper, we investigate the promising
idea of using large language models (LLMs) to stem words by leveraging its
capability of context understanding. With this respect, we identify three
avenues, each characterised by different trade-offs in terms of computational
cost, effectiveness and robustness : (1) use LLMs to stem the vocabulary for a
collection, i.e., the set of unique words that appear in the collection
(vocabulary stemming), (2) use LLMs to stem each document separately
(contextual stemming), and (3) use LLMs to extract from each document entities
that should not be stemmed, then use vocabulary stemming to stem the rest of
the terms (entity-based contextual stemming). Through a series of empirical
experiments, we compare the use of LLMs for stemming with that of traditional
lexical stemmers such as Porter and Krovetz for English text. We find that
while vocabulary stemming and contextual stemming fail to achieve higher
effectiveness than traditional stemmers, entity-based contextual stemming can
achieve a higher effectiveness than using Porter stemmer alone, under specific
conditions. | cs.IR | Large Language Models for Stemming: Promises, Pitfalls and Failures |
2024-02-19T01:04:22Z | http://arxiv.org/pdf/2402.11756v3 | http://arxiv.org/abs/2402.11756v3 | Yavuz Faruk Bakman, Duygu Nur Yaldiz, Baturalp Buyukates, Chenyang Tao, Dimitrios Dimitriadis, Salman Avestimehr | 2024-06-08T20:40:55Z | cs.CL, cs.LG | Generative Large Language Models (LLMs) are widely utilized for their
excellence in various tasks. However, their tendency to produce inaccurate or
misleading outputs poses a potential risk, particularly in high-stakes
environments. Therefore, estimating the correctness of generative LLM outputs
is an important task for enhanced reliability. Uncertainty Estimation (UE) in
generative LLMs is an evolving domain, where SOTA probability-based methods
commonly employ length-normalized scoring. In this work, we propose
Meaning-Aware Response Scoring (MARS) as an alternative to length-normalized
scoring for UE methods. MARS is a novel scoring function that considers the
semantic contribution of each token in the generated sequence in the context of
the question. We demonstrate that integrating MARS into UE methods results in a
universal and significant improvement in UE performance. We conduct experiments
using three distinct closed-book question-answering datasets across five
popular pre-trained LLMs. Lastly, we validate the efficacy of MARS on a Medical
QA dataset. Code can be found https://github.com/Ybakman/LLM_Uncertainity. | cs.CL | MARS: Meaning-Aware Response Scoring for Uncertainty Estimation in Generative LLMs |
2024-02-19T00:53:48Z | http://arxiv.org/pdf/2402.11755v1 | http://arxiv.org/abs/2402.11755v1 | Reshabh K Sharma, Vinayak Gupta, Dan Grossman | 2024-02-19T00:53:48Z | cs.LG, cs.CL, cs.CR, cs.PL | Large language models (LLMs) have profoundly transformed natural language
applications, with a growing reliance on instruction-based definitions for
designing chatbots. However, post-deployment the chatbot definitions are fixed
and are vulnerable to attacks by malicious users, emphasizing the need to
prevent unethical applications and financial losses. Existing studies explore
user prompts' impact on LLM-based chatbots, yet practical methods to contain
attacks on application-specific chatbots remain unexplored. This paper presents
System Prompt Meta Language (SPML), a domain-specific language for refining
prompts and monitoring the inputs to the LLM-based chatbots. SPML actively
checks attack prompts, ensuring user inputs align with chatbot definitions to
prevent malicious execution on the LLM backbone, optimizing costs. It also
streamlines chatbot definition crafting with programming language capabilities,
overcoming natural language design challenges. Additionally, we introduce a
groundbreaking benchmark with 1.8k system prompts and 20k user inputs, offering
the inaugural language and benchmark for chatbot definition evaluation.
Experiments across datasets demonstrate SPML's proficiency in understanding
attacker prompts, surpassing models like GPT-4, GPT-3.5, and LLAMA. Our data
and codes are publicly available at: https://prompt-compiler.github.io/SPML/. | cs.LG | SPML: A DSL for Defending Language Models Against Prompt Attacks |
2024-02-19T00:43:31Z | http://arxiv.org/pdf/2402.11753v4 | http://arxiv.org/abs/2402.11753v4 | Fengqing Jiang, Zhangchen Xu, Luyao Niu, Zhen Xiang, Bhaskar Ramasubramanian, Bo Li, Radha Poovendran | 2024-06-07T17:35:17Z | cs.CL, cs.AI | Safety is critical to the usage of large language models (LLMs). Multiple
techniques such as data filtering and supervised fine-tuning have been
developed to strengthen LLM safety. However, currently known techniques presume
that corpora used for safety alignment of LLMs are solely interpreted by
semantics. This assumption, however, does not hold in real-world applications,
which leads to severe vulnerabilities in LLMs. For example, users of forums
often use ASCII art, a form of text-based art, to convey image information. In
this paper, we propose a novel ASCII art-based jailbreak attack and introduce a
comprehensive benchmark Vision-in-Text Challenge (ViTC) to evaluate the
capabilities of LLMs in recognizing prompts that cannot be solely interpreted
by semantics. We show that five SOTA LLMs (GPT-3.5, GPT-4, Gemini, Claude, and
Llama2) struggle to recognize prompts provided in the form of ASCII art. Based
on this observation, we develop the jailbreak attack ArtPrompt, which leverages
the poor performance of LLMs in recognizing ASCII art to bypass safety measures
and elicit undesired behaviors from LLMs. ArtPrompt only requires black-box
access to the victim LLMs, making it a practical attack. We evaluate ArtPrompt
on five SOTA LLMs, and show that ArtPrompt can effectively and efficiently
induce undesired behaviors from all five LLMs. Our code is available at
https://github.com/uw-nsl/ArtPrompt. | cs.CL | ArtPrompt: ASCII Art-based Jailbreak Attacks against Aligned LLMs |
2024-02-19T00:40:17Z | http://arxiv.org/pdf/2402.14838v1 | http://arxiv.org/abs/2402.14838v1 | Mohammad Heydari Rad, Farhan Farsi, Shayan Bali, Romina Etezadi, Mehrnoush Shamsfard | 2024-02-19T00:40:17Z | cs.CL, cs.AI, cs.LG | Nowadays, the usage of Large Language Models (LLMs) has increased, and LLMs
have been used to generate texts in different languages and for different
tasks. Additionally, due to the participation of remarkable companies such as
Google and OpenAI, LLMs are now more accessible, and people can easily use
them. However, an important issue is how we can detect AI-generated texts from
human-written ones. In this article, we have investigated the problem of
AI-generated text detection from two different aspects: semantics and syntax.
Finally, we presented an AI model that can distinguish AI-generated texts from
human-written ones with high accuracy on both multilingual and monolingual
tasks using the M4 dataset. According to our results, using a semantic approach
would be more helpful for detection. However, there is a lot of room for
improvement in the syntactic approach, and it would be a good approach for
future work. | cs.CL | RFBES at SemEval-2024 Task 8: Investigating Syntactic and Semantic Features for Distinguishing AI-Generated and Human-Written Texts |
2024-02-19T00:39:31Z | http://arxiv.org/pdf/2402.11750v2 | http://arxiv.org/abs/2402.11750v2 | Vinay M. S., Minh-Hao Van, Xintao Wu | 2024-06-17T18:34:54Z | cs.CL | Large Language Models (LLMs) have showcased their In-Context Learning (ICL)
capabilities, enabling few-shot learning without the need for gradient updates.
Despite its advantages, the effectiveness of ICL heavily depends on the choice
of demonstrations. Selecting the most effective demonstrations for ICL remains
a significant research challenge. To tackle this issue, we propose a
demonstration selection method named InfICL, which utilizes influence functions
to analyze impacts of training samples. By identifying the most influential
training samples as demonstrations, InfICL aims to enhance the ICL
generalization performance. To keep InfICL cost-effective, we only use the LLM
to generate sample input embeddings, avoiding expensive fine-tuning. Through
empirical studies on various real-world datasets, we demonstrate advantages of
InfICL compared to state-of-the-art baselines. | cs.CL | In-Context Learning Demonstration Selection via Influence Analysis |
2024-02-19T00:18:09Z | http://arxiv.org/pdf/2402.11746v1 | http://arxiv.org/abs/2402.11746v1 | Rishabh Bhardwaj, Do Duc Anh, Soujanya Poria | 2024-02-19T00:18:09Z | cs.CL, cs.AI | Aligned language models face a significant limitation as their fine-tuning
often results in compromised safety. To tackle this, we propose a simple method
RESTA that performs LLM safety realignment. RESTA stands for REstoring Safety
through Task Arithmetic. At its core, it involves a simple arithmetic addition
of a safety vector to the weights of the compromised model. We demonstrate the
effectiveness of RESTA in both parameter-efficient and full fine-tuning,
covering a wide range of downstream tasks, including instruction following in
Chinese, English, and Hindi, as well as problem-solving capabilities in Code
and Math. We also showcase the generalizability of RESTA on three existing
safety evaluation benchmarks and a multilingual benchmark dataset proposed as a
part of this work, consisting of 550 harmful questions covering 11 categories,
each with 5 sub-categories of harm. Overall, RESTA decreases the harmfulness of
the compromised model from 18.6% to 5.1% and from 9.2% to 1.5% in
parameter-efficient and full fine-tuning, respectively, while maintaining most
of the model's performance on the task. We release the source codes at:
https://github.com/declare-lab/resta. | cs.CL | Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmetic |
2024-02-19T00:07:28Z | http://arxiv.org/pdf/2402.11744v2 | http://arxiv.org/abs/2402.11744v2 | Zhongping Zhang, Wenda Qin, Bryan A. Plummer | 2024-06-10T19:20:20Z | cs.CL | Machine-Generated Text (MGT) detection aims to identify a piece of text as
machine or human written. Prior work has primarily formulated MGT detection as
a binary classification task over an entire document, with limited work
exploring cases where only part of a document is machine generated. This paper
provides the first in-depth study of MGT that localizes the portions of a
document that were machine generated. Thus, if a bad actor were to change a key
portion of a news article to spread misinformation, whole document MGT
detection may fail since the vast majority is human written, but our approach
can succeed due to its granular approach. A key challenge in our MGT
localization task is that short spans of text, e.g., a single sentence,
provides little information indicating if it is machine generated due to its
short length. To address this, we leverage contextual information, where we
predict whether multiple sentences are machine or human written at once. This
enables our approach to identify changes in style or content to boost
performance. A gain of 4-13% mean Average Precision (mAP) over prior work
demonstrates the effectiveness of approach on five diverse datasets: GoodNews,
VisualNews, WikiText, Essay, and WP. We release our implementation at
https://github.com/Zhongping-Zhang/MGT_Localization. | cs.CL | Machine-Generated Text Localization |
2024-02-18T23:22:40Z | http://arxiv.org/pdf/2403.00784v1 | http://arxiv.org/abs/2403.00784v1 | Jiajia Wang, Jimmy X. Huang, Xinhui Tu, Junmei Wang, Angela J. Huang, Md Tahmid Rahman Laskar, Amran Bhuiyan | 2024-02-18T23:22:40Z | cs.IR, cs.AI, cs.CL | Recent years have witnessed a substantial increase in the use of deep
learning to solve various natural language processing (NLP) problems. Early
deep learning models were constrained by their sequential or unidirectional
nature, such that they struggled to capture the contextual relationships across
text inputs. The introduction of bidirectional encoder representations from
transformers (BERT) leads to a robust encoder for the transformer model that
can understand the broader context and deliver state-of-the-art performance
across various NLP tasks. This has inspired researchers and practitioners to
apply BERT to practical problems, such as information retrieval (IR). A survey
that focuses on a comprehensive analysis of prevalent approaches that apply
pretrained transformer encoders like BERT to IR can thus be useful for academia
and the industry. In light of this, we revisit a variety of BERT-based methods
in this survey, cover a wide range of techniques of IR, and group them into six
high-level categories: (i) handling long documents, (ii) integrating semantic
information, (iii) balancing effectiveness and efficiency, (iv) predicting the
weights of terms, (v) query expansion, and (vi) document expansion. We also
provide links to resources, including datasets and toolkits, for BERT-based IR
systems. A key highlight of our survey is the comparison between BERT's
encoder-based models and the latest generative Large Language Models (LLMs),
such as ChatGPT, which rely on decoders. Despite the popularity of LLMs, we
find that for specific tasks, finely tuned BERT encoders still outperform, and
at a lower deployment cost. Finally, we summarize the comprehensive outcomes of
the survey and suggest directions for future research in the area. | cs.IR | Utilizing BERT for Information Retrieval: Survey, Applications, Resources, and Challenges |
2024-02-18T23:03:56Z | http://arxiv.org/pdf/2402.14837v1 | http://arxiv.org/abs/2402.14837v1 | Oluwole Fagbohun, Rachel M. Harrison, Anton Dereventsov | 2024-02-18T23:03:56Z | cs.CL, cs.AI, cs.HC, cs.LG | Due to rapid advancements in the development of Large Language Models (LLMs),
programming these models with prompts has recently gained significant
attention. However, the sheer number of available prompt engineering techniques
creates an overwhelming landscape for practitioners looking to utilize these
tools. For the most efficient and effective use of LLMs, it is important to
compile a comprehensive list of prompting techniques and establish a
standardized, interdisciplinary categorization framework. In this survey, we
examine some of the most well-known prompting techniques from both academic and
practical viewpoints and classify them into seven distinct categories. We
present an overview of each category, aiming to clarify their unique
contributions and showcase their practical applications in real-world examples
in order to equip fellow practitioners with a structured framework for
understanding and categorizing prompting techniques tailored to their specific
domains. We believe that this approach will help simplify the complex landscape
of prompt engineering and enable more effective utilization of LLMs in various
applications. By providing practitioners with a systematic approach to prompt
categorization, we aim to assist in navigating the intricacies of effective
prompt design for conversational pre-trained LLMs and inspire new possibilities
in their respective fields. | cs.CL | An Empirical Categorization of Prompting Techniques for Large Language Models: A Practitioner's Guide |
2024-02-18T22:55:26Z | http://arxiv.org/pdf/2402.11728v1 | http://arxiv.org/abs/2402.11728v1 | Agam Shah, Arnav Hiray, Pratvi Shah, Arkaprabha Banerjee, Anushka Singh, Dheeraj Eidnani, Bhaskar Chaudhury, Sudheer Chava | 2024-02-18T22:55:26Z | cs.CL, cs.LG, q-fin.CP | In this paper, we investigate the influence of claims in analyst reports and
earnings calls on financial market returns, considering them as significant
quarterly events for publicly traded companies. To facilitate a comprehensive
analysis, we construct a new financial dataset for the claim detection task in
the financial domain. We benchmark various language models on this dataset and
propose a novel weak-supervision model that incorporates the knowledge of
subject matter experts (SMEs) in the aggregation function, outperforming
existing approaches. Furthermore, we demonstrate the practical utility of our
proposed model by constructing a novel measure ``optimism". Furthermore, we
observed the dependence of earnings surprise and return on our optimism
measure. Our dataset, models, and code will be made publicly (under CC BY 4.0
license) available on GitHub and Hugging Face. | cs.CL | Numerical Claim Detection in Finance: A New Financial Dataset, Weak-Supervision Model, and Market Analysis |
2024-02-18T22:36:19Z | http://arxiv.org/pdf/2402.11725v3 | http://arxiv.org/abs/2402.11725v3 | Kai Chen, Zihao He, Jun Yan, Taiwei Shi, Kristina Lerman | 2024-06-18T05:14:02Z | cs.CL, cs.CR, cs.CY | Large Language Models (LLMs) possess the potential to exert substantial
influence on public perceptions and interactions with information. This raises
concerns about the societal impact that could arise if the ideologies within
these models can be easily manipulated. In this work, we investigate how
effectively LLMs can learn and generalize ideological biases from their
instruction-tuning data. Our findings reveal a concerning vulnerability:
exposure to only a small amount of ideologically driven samples significantly
alters the ideology of LLMs. Notably, LLMs demonstrate a startling ability to
absorb ideology from one topic and generalize it to even unrelated ones. The
ease with which LLMs' ideologies can be skewed underscores the risks associated
with intentionally poisoned training data by malicious actors or inadvertently
introduced biases by data annotators. It also emphasizes the imperative for
robust safeguards to mitigate the influence of ideological manipulations on
LLMs. | cs.CL | How Susceptible are Large Language Models to Ideological Manipulation? |
2024-02-18T22:27:42Z | http://arxiv.org/pdf/2402.11723v1 | http://arxiv.org/abs/2402.11723v1 | Paramveer S. Dhillon, Somayeh Molaei, Jiaqi Li, Maximilian Golub, Shaochun Zheng, Lionel P. Robert | 2024-02-18T22:27:42Z | cs.HC, cs.CL | Advances in language modeling have paved the way for novel human-AI
co-writing experiences. This paper explores how varying levels of scaffolding
from large language models (LLMs) shape the co-writing process. Employing a
within-subjects field experiment with a Latin square design, we asked
participants (N=131) to respond to argumentative writing prompts under three
randomly sequenced conditions: no AI assistance (control), next-sentence
suggestions (low scaffolding), and next-paragraph suggestions (high
scaffolding). Our findings reveal a U-shaped impact of scaffolding on writing
quality and productivity (words/time). While low scaffolding did not
significantly improve writing quality or productivity, high scaffolding led to
significant improvements, especially benefiting non-regular writers and less
tech-savvy users. No significant cognitive burden was observed while using the
scaffolded writing tools, but a moderate decrease in text ownership and
satisfaction was noted. Our results have broad implications for the design of
AI-powered writing tools, including the need for personalized scaffolding
mechanisms. | cs.HC | Shaping Human-AI Collaboration: Varied Scaffolding Levels in Co-writing with Language Models |
2024-02-18T21:28:06Z | http://arxiv.org/pdf/2402.11712v1 | http://arxiv.org/abs/2402.11712v1 | Farhad Moghimifar, Yuan-Fang Li, Robert Thomson, Gholamreza Haffari | 2024-02-18T21:28:06Z | cs.CL | Coalition negotiations are a cornerstone of parliamentary democracies,
characterised by complex interactions and strategic communications among
political parties. Despite its significance, the modelling of these
negotiations has remained unexplored with the domain of Natural Language
Processing (NLP), mostly due to lack of proper data. In this paper, we
introduce coalition negotiations as a novel NLP task, and model it as a
negotiation between large language model-based agents. We introduce a
multilingual dataset, POLCA, comprising manifestos of European political
parties and coalition agreements over a number of elections in these countries.
This dataset addresses the challenge of the current scope limitations in
political negotiation modelling by providing a diverse, real-world basis for
simulation. Additionally, we propose a hierarchical Markov decision process
designed to simulate the process of coalition negotiation between political
parties and predict the outcomes. We evaluate the performance of
state-of-the-art large language models (LLMs) as agents in handling coalition
negotiations, offering insights into their capabilities and paving the way for
future advancements in political modelling. | cs.CL | Modelling Political Coalition Negotiations Using LLM-based Agents |
2024-02-18T21:25:09Z | http://arxiv.org/pdf/2402.11711v1 | http://arxiv.org/abs/2402.11711v1 | Yasaman Jafari, Dheeraj Mekala, Rose Yu, Taylor Berg-Kirkpatrick | 2024-02-18T21:25:09Z | cs.CL | RL-based techniques can be used to search for prompts that when fed into a
target language model maximize a set of user-specified reward functions.
However, in many target applications, the natural reward functions are in
tension with one another -- for example, content preservation vs. style
matching in style transfer tasks. Current techniques focus on maximizing the
average of reward functions, which does not necessarily lead to prompts that
achieve balance across rewards -- an issue that has been well-studied in the
multi-objective and robust optimization literature. In this paper, we adapt
several techniques for multi-objective optimization to RL-based discrete prompt
optimization -- two that consider volume of the Pareto reward surface, and
another that chooses an update direction that benefits all rewards
simultaneously. We conduct an empirical analysis of these methods on two NLP
tasks: style transfer and machine translation, each using three competing
reward functions. Our experiments demonstrate that multi-objective methods that
directly optimize volume perform better and achieve a better balance of all
rewards than those that attempt to find monotonic update directions. | cs.CL | MORL-Prompt: An Empirical Analysis of Multi-Objective Reinforcement Learning for Discrete Prompt Optimization |
2024-02-18T21:20:33Z | http://arxiv.org/pdf/2402.11710v1 | http://arxiv.org/abs/2402.11710v1 | Jia Xu, Mona Diab | 2024-02-18T21:20:33Z | cs.CL | Minimizing social bias strengthens societal bonds, promoting shared
understanding and better decision-making. We revisit the definition of bias by
discovering new bias types (e.g., societal status) in dynamic environments and
describe them relative to context, such as culture, region, time, and personal
background. Our framework includes eight hypotheses about bias and a minimizing
bias strategy for each assumption as well as five methods as proposed solutions
in LLM. The realization of the framework is yet to be completed. | cs.CL | A Note on Bias to Complete |
2024-02-18T21:13:05Z | http://arxiv.org/pdf/2402.11709v2 | http://arxiv.org/abs/2402.11709v2 | Shuzhou Yuan, Ercong Nie, Michael Färber, Helmut Schmid, Hinrich Schütze | 2024-06-07T14:36:33Z | cs.CL, cs.AI | Large Language Models (LLMs) exhibit strong In-Context Learning (ICL)
capabilities when prompts with demonstrations are used. However, fine-tuning
still remains crucial to further enhance their adaptability. Prompt-based
fine-tuning proves to be an effective fine-tuning method in low-data scenarios,
but high demands on computing resources limit its practicality. We address this
issue by introducing a prompt-based parameter-efficient fine-tuning (PEFT)
approach. GNNavi leverages insights into ICL's information flow dynamics, which
indicates that label words act in prompts as anchors for information
propagation. GNNavi employs a Graph Neural Network (GNN) layer to precisely
guide the aggregation and distribution of information flow during the
processing of prompts by hardwiring the desired information flow into the GNN.
Our experiments on text classification tasks with GPT-2 and Llama2 show GNNavi
surpasses standard prompt-based fine-tuning methods in few-shot settings by
updating just 0.2% to 0.5% of parameters. We compare GNNavi with prevalent PEFT
approaches, such as prefix tuning, LoRA and Adapter in terms of performance and
efficiency. Our analysis reveals that GNNavi enhances information flow and
ensures a clear aggregation process. | cs.CL | GNNavi: Navigating the Information Flow in Large Language Models by Graph Neural Network |
2024-02-18T20:47:10Z | http://arxiv.org/pdf/2402.11700v1 | http://arxiv.org/abs/2402.11700v1 | Shuzhou Yuan, Ercong Nie, Bolei Ma, Michael Färber | 2024-02-18T20:47:10Z | cs.CL | Large Language Models (LLMs) possess outstanding capabilities in addressing
various natural language processing (NLP) tasks. However, the sheer size of
these models poses challenges in terms of storage, training and inference due
to the inclusion of billions of parameters through layer stacking. While
traditional approaches such as model pruning or distillation offer ways for
reducing model size, they often come at the expense of performance retention.
In our investigation, we systematically explore the approach of reducing the
number of layers in LLMs. Surprisingly, we observe that even with fewer layers,
LLMs maintain similar or better performance levels, particularly in
prompt-based fine-tuning for text classification tasks. Remarkably, in certain
cases, models with a single layer outperform their fully layered counterparts.
These findings offer valuable insights for future work aimed at mitigating the
size constraints of LLMs while preserving their performance, thereby opening
avenues for significantly more efficient use of LLMs. | cs.CL | Why Lift so Heavy? Slimming Large Language Models by Cutting Off the Layers |
2024-02-18T19:38:44Z | http://arxiv.org/pdf/2402.11690v1 | http://arxiv.org/abs/2402.11690v1 | Zhiyang Xu, Chao Feng, Rulin Shao, Trevor Ashby, Ying Shen, Di Jin, Yu Cheng, Qifan Wang, Lifu Huang | 2024-02-18T19:38:44Z | cs.CL, cs.CV | Despite vision-language models' (VLMs) remarkable capabilities as versatile
visual assistants, two substantial challenges persist within the existing VLM
frameworks: (1) lacking task diversity in pretraining and visual instruction
tuning, and (2) annotation error and bias in GPT-4 synthesized instruction
tuning data. Both challenges lead to issues such as poor generalizability,
hallucination, and catastrophic forgetting. To address these challenges, we
construct Vision-Flan, the most diverse publicly available visual instruction
tuning dataset to date, comprising 187 diverse tasks and 1,664,261 instances
sourced from academic datasets, and each task is accompanied by an
expert-written instruction. In addition, we propose a two-stage instruction
tuning framework, in which VLMs are firstly finetuned on Vision-Flan and
further tuned on GPT-4 synthesized data. We find this two-stage tuning
framework significantly outperforms the traditional single-stage visual
instruction tuning framework and achieves the state-of-the-art performance
across a wide range of multi-modal evaluation benchmarks. Finally, we conduct
in-depth analyses to understand visual instruction tuning and our findings
reveal that: (1) GPT-4 synthesized data does not substantially enhance VLMs'
capabilities but rather modulates the model's responses to human-preferred
formats; (2) A minimal quantity (e.g., 1,000) of GPT-4 synthesized data can
effectively align VLM responses with human-preference; (3) Visual instruction
tuning mainly helps large-language models (LLMs) to understand visual features. | cs.CL | Vision-Flan: Scaling Human-Labeled Tasks in Visual Instruction Tuning |
2024-02-18T19:26:49Z | http://arxiv.org/pdf/2402.11684v2 | http://arxiv.org/abs/2402.11684v2 | Guiming Hardy Chen, Shunian Chen, Ruifei Zhang, Junying Chen, Xiangbo Wu, Zhiyi Zhang, Zhihong Chen, Jianquan Li, Xiang Wan, Benyou Wang | 2024-06-17T07:55:59Z | cs.CL, cs.AI | Large vision-language models (LVLMs) have shown premise in a broad range of
vision-language tasks with their strong reasoning and generalization
capabilities. However, they require considerable computational resources for
training and deployment. This study aims to bridge the performance gap between
traditional-scale LVLMs and resource-friendly lite versions by adopting
high-quality training data. To this end, we propose a comprehensive pipeline
for generating a synthetic dataset. The key idea is to leverage strong
proprietary models to generate (i) fine-grained image annotations for
vision-language alignment and (ii) complex reasoning visual question-answering
pairs for visual instruction fine-tuning, yielding 1.3M samples in total. We
train a series of lite VLMs on the synthetic dataset and experimental results
demonstrate the effectiveness of the proposed scheme, where they achieve
competitive performance on 17 benchmarks among 4B LVLMs, and even perform on
par with 7B/13B-scale models on various benchmarks. This work highlights the
feasibility of adopting high-quality data in crafting more efficient LVLMs. We
name our dataset \textit{ALLaVA}, and open-source it to research community for
developing better resource-efficient LVLMs for wider usage. | cs.CL | ALLaVA: Harnessing GPT4V-Synthesized Data for Lite Vision-Language Models |
2024-02-18T19:13:52Z | http://arxiv.org/pdf/2402.11683v2 | http://arxiv.org/abs/2402.11683v2 | Tejpalsingh Siledar, Swaroop Nath, Sankara Sri Raghava Ravindra Muddu, Rupasai Rangaraju, Swaprava Nath, Pushpak Bhattacharyya, Suman Banerjee, Amey Patil, Sudhanshu Shekhar Singh, Muthusamy Chelliah, Nikesh Garera | 2024-06-09T06:10:22Z | cs.CL | Evaluation of opinion summaries using conventional reference-based metrics
rarely provides a holistic evaluation and has been shown to have a relatively
low correlation with human judgments. Recent studies suggest using Large
Language Models (LLMs) as reference-free metrics for NLG evaluation, however,
they remain unexplored for opinion summary evaluation. Moreover, limited
opinion summary evaluation datasets inhibit progress. To address this, we
release the SUMMEVAL-OP dataset covering 7 dimensions related to the evaluation
of opinion summaries: fluency, coherence, relevance, faithfulness, aspect
coverage, sentiment consistency, and specificity. We investigate Op-I-Prompt a
dimension-independent prompt, and Op-Prompts, a dimension-dependent set of
prompts for opinion summary evaluation. Experiments indicate that Op-I-Prompt
emerges as a good alternative for evaluating opinion summaries achieving an
average Spearman correlation of 0.70 with humans, outperforming all previous
approaches. To the best of our knowledge, we are the first to investigate LLMs
as evaluators on both closed-source and open-source models in the opinion
summarization domain. | cs.CL | One Prompt To Rule Them All: LLMs for Opinion Summary Evaluation |
2024-02-18T19:11:58Z | http://arxiv.org/pdf/2402.11681v1 | http://arxiv.org/abs/2402.11681v1 | Jérôme Michaud, Anna Jon-and | 2024-02-18T19:11:58Z | cs.CL, cs.NA, math.NA | Recent advances in large language models using deep learning techniques have
renewed interest on how languages can be learned from data. However, it is
unclear whether or how these models represent grammatical information from the
learned languages. In addition, the models must be pre-trained on large corpora
before they can be used. In this work, we propose an alternative, more
transparent and cognitively plausible architecture for learning language.
Instead of using deep learning, our approach uses a minimal cognitive
architecture based on sequence memory and chunking. The learning mechanism is
based on the principles of reinforcement learning. We test our architecture on
a number of natural-like toy languages. Results show that the model can learn
these artificial languages from scratch and extract grammatical information
that supports learning. Our study demonstrates the power of this simple
architecture and stresses the importance of sequence memory as a key component
of the language learning process. Since other animals do not seem to have a
faithful sequence memory, this may explain why only humans have developed
complex languages. | cs.CL | Opening the black box of language acquisition |
2024-02-18T18:56:07Z | http://arxiv.org/pdf/2402.11676v2 | http://arxiv.org/abs/2402.11676v2 | Jaylen Jones, Lingbo Mo, Eric Fosler-Lussier, Huan Sun | 2024-03-29T15:01:38Z | cs.CL, cs.AI | Counter narratives - informed responses to hate speech contexts designed to
refute hateful claims and de-escalate encounters - have emerged as an effective
hate speech intervention strategy. While previous work has proposed automatic
counter narrative generation methods to aid manual interventions, the
evaluation of these approaches remains underdeveloped. Previous automatic
metrics for counter narrative evaluation lack alignment with human judgment as
they rely on superficial reference comparisons instead of incorporating key
aspects of counter narrative quality as evaluation criteria. To address prior
evaluation limitations, we propose a novel evaluation framework prompting LLMs
to provide scores and feedback for generated counter narrative candidates using
5 defined aspects derived from guidelines from counter narrative specialized
NGOs. We found that LLM evaluators achieve strong alignment to human-annotated
scores and feedback and outperform alternative metrics, indicating their
potential as multi-aspect, reference-free and interpretable evaluators for
counter narrative evaluation. | cs.CL | A Multi-Aspect Framework for Counter Narrative Evaluation using Large Language Models |
2024-02-18T18:20:57Z | http://arxiv.org/pdf/2402.11671v1 | http://arxiv.org/abs/2402.11671v1 | Agnes Luhtaru, Martin Vainikko, Krista Liin, Kais Allkivi-Metsoja, Jaagup Kippar, Pille Eslon, Mark Fishel | 2024-02-18T18:20:57Z | cs.CL, cs.AI | The project was funded in 2021-2023 by the National Programme of Estonian
Language Technology. Its main aim was to develop spelling and grammar
correction tools for the Estonian language. The main challenge was the very
small amount of available error correction data needed for such development. To
mitigate this, (1) we annotated more correction data for model training and
testing, (2) we tested transfer-learning, i.e. retraining machine learning
models created for other tasks, so as not to depend solely on correction data,
(3) we compared the developed method and model with alternatives, including
large language models. We also developed automatic evaluation, which can
calculate the accuracy and yield of corrections by error category, so that the
effectiveness of different methods can be compared in detail.
There has been a breakthrough in large language models during the project:
GPT4, a commercial language model with Estonian-language support, has been
created. We took into account the existence of the model when adjusting plans
and in the report we present a comparison with the ability of GPT4 to improve
the Estonian language text.
The final results show that the approach we have developed provides better
scores than GPT4 and the result is usable but not entirely reliable yet. The
report also contains ideas on how GPT4 and other major language models can be
implemented in the future, focusing on open-source solutions.
All results of this project are open-data/open-source, with licenses that
allow them to be used for purposes including commercial ones. | cs.CL | Autocorrect for Estonian texts: final report from project EKTB25 |
2024-02-18T17:27:51Z | http://arxiv.org/pdf/2402.11656v2 | http://arxiv.org/abs/2402.11656v2 | Ju-Hyung Lee, Dong-Ho Lee, Joohan Lee, Jay Pujara | 2024-06-28T23:00:45Z | cs.IT, cs.CL, cs.LG, eess.SP, math.IT | The burgeoning field of on-device AI communication, where devices exchange
information directly through embedded foundation models, such as language
models (LMs), requires robust, efficient, and generalizable communication
frameworks. However, integrating these frameworks with existing wireless
systems and effectively managing noise and bit errors pose significant
challenges. In this work, we introduce a practical ondevice AI communication
framework, integrated with physical layer (PHY) communication functions,
demonstrated through its performance on a link-level simulator. Our framework
incorporates end-to-end training with channel noise to enhance resilience,
incorporates vector quantized variational autoencoders (VQ-VAE) for efficient
and robust communication, and utilizes pre-trained encoder-decoder transformers
for improved generalization capabilities. Simulations, across various
communication scenarios, reveal that our framework achieves a 50% reduction in
transmission size while demonstrating substantial generalization ability and
noise robustness under standardized 3GPP channel models. | cs.IT | Integrating Pre-Trained Language Model with Physical Layer Communications |
2024-02-18T17:26:51Z | http://arxiv.org/pdf/2402.11655v2 | http://arxiv.org/abs/2402.11655v2 | Francesco Ortu, Zhijing Jin, Diego Doimo, Mrinmaya Sachan, Alberto Cazzaniga, Bernhard Schölkopf | 2024-06-06T21:45:21Z | cs.CL | Interpretability research aims to bridge the gap between empirical success
and our scientific understanding of the inner workings of large language models
(LLMs). However, most existing research focuses on analyzing a single
mechanism, such as how models copy or recall factual knowledge. In this work,
we propose a formulation of competition of mechanisms, which focuses on the
interplay of multiple mechanisms instead of individual mechanisms and traces
how one of them becomes dominant in the final prediction. We uncover how and
where mechanisms compete within LLMs using two interpretability methods: logit
inspection and attention modification. Our findings show traces of the
mechanisms and their competition across various model components and reveal
attention positions that effectively control the strength of certain
mechanisms. Code: https://github.com/francescortu/comp-mech. Data:
https://huggingface.co/datasets/francescortu/comp-mech. | cs.CL | Competition of Mechanisms: Tracing How Language Models Handle Facts and Counterfactuals |
2024-02-18T17:10:07Z | http://arxiv.org/pdf/2402.11651v2 | http://arxiv.org/abs/2402.11651v2 | Renxi Wang, Haonan Li, Xudong Han, Yixuan Zhang, Timothy Baldwin | 2024-04-16T11:41:13Z | cs.CL, I.2.7 | Large language models (LLMs) have achieved success in acting as agents, which
interact with environments through tools such as search engines. However, LLMs
are optimized for language generation instead of tool use during training or
alignment, limiting their effectiveness as agents. To resolve this problem,
previous work has first collected interaction trajectories between LLMs and
environments, using only trajectories that successfully finished the task to
fine-tune smaller models, making fine-tuning data scarce and acquiring it both
difficult and costly. Discarding failed trajectories also leads to significant
wastage of data and resources and limits the possible optimization paths during
fine-tuning. In this paper, we argue that unsuccessful trajectories offer
valuable insights, and LLMs can learn from these trajectories through
appropriate quality control and fine-tuning strategies. By simply adding a
prefix or suffix that tells the model whether to generate a successful
trajectory during training, we improve model performance by a large margin on
mathematical reasoning, multi-hop question answering, and strategic question
answering tasks. We further analyze the inference results and find that our
method provides a better trade-off between valuable information and errors in
unsuccessful trajectories. To our knowledge, we are the first to demonstrate
the value of negative trajectories and their application in agent-tunning
scenarios. Our findings offer guidance for developing better agent-tuning
methods and low-resource data usage techniques. | cs.CL | Learning From Failure: Integrating Negative Examples when Fine-tuning Large Language Models as Agents |
2024-02-18T16:51:02Z | http://arxiv.org/pdf/2402.14836v2 | http://arxiv.org/abs/2402.14836v2 | Jinghao Zhang, Yuting Liu, Qiang Liu, Shu Wu, Guibing Guo, Liang Wang | 2024-06-05T17:02:47Z | cs.CL, cs.AI, cs.IR | Recently, the powerful large language models (LLMs) have been instrumental in
propelling the progress of recommender systems (RS). However, while these
systems have flourished, their susceptibility to security threats has been
largely overlooked. In this work, we reveal that the introduction of LLMs into
recommendation models presents new security vulnerabilities due to their
emphasis on the textual content of items. We demonstrate that attackers can
significantly boost an item's exposure by merely altering its textual content
during the testing phase, without requiring direct interference with the
model's training process. Additionally, the attack is notably stealthy, as it
does not affect the overall recommendation performance and the modifications to
the text are subtle, making it difficult for users and platforms to detect. Our
comprehensive experiments across four mainstream LLM-based recommendation
models demonstrate the superior efficacy and stealthiness of our approach. Our
work unveils a significant security gap in LLM-based recommendation systems and
paves the way for future research on protecting these systems. | cs.CL | Stealthy Attack on Large Language Model based Recommendation |
2024-02-18T16:37:32Z | http://arxiv.org/pdf/2402.11639v2 | http://arxiv.org/abs/2402.11639v2 | Liam Collins, Advait Parulekar, Aryan Mokhtari, Sujay Sanghavi, Sanjay Shakkottai | 2024-05-28T05:15:53Z | cs.LG, cs.AI, cs.CL | A striking property of transformers is their ability to perform in-context
learning (ICL), a machine learning framework in which the learner is presented
with a novel context during inference implicitly through some data, and tasked
with making a prediction in that context. As such, that learner must adapt to
the context without additional training. We explore the role of softmax
attention in an ICL setting where each context encodes a regression task. We
show that an attention unit learns a window that it uses to implement a
nearest-neighbors predictor adapted to the landscape of the pretraining tasks.
Specifically, we show that this window widens with decreasing Lipschitzness and
increasing label noise in the pretraining tasks. We also show that on low-rank,
linear problems, the attention unit learns to project onto the appropriate
subspace before inference. Further, we show that this adaptivity relies
crucially on the softmax activation and thus cannot be replicated by the linear
activation often studied in prior theoretical analyses. | cs.LG | In-Context Learning with Transformers: Softmax Attention Adapts to Function Lipschitzness |
2024-02-18T16:36:00Z | http://arxiv.org/pdf/2402.11638v1 | http://arxiv.org/abs/2402.11638v1 | Yichen Wang, Shangbin Feng, Abe Bohan Hou, Xiao Pu, Chao Shen, Xiaoming Liu, Yulia Tsvetkov, Tianxing He | 2024-02-18T16:36:00Z | cs.CL | The widespread use of large language models (LLMs) is increasing the demand
for methods that detect machine-generated text to prevent misuse. The goal of
our study is to stress test the detectors' robustness to malicious attacks
under realistic scenarios. We comprehensively study the robustness of popular
machine-generated text detectors under attacks from diverse categories:
editing, paraphrasing, prompting, and co-generating. Our attacks assume limited
access to the generator LLMs, and we compare the performance of detectors on
different attacks under different budget levels. Our experiments reveal that
almost none of the existing detectors remain robust under all the attacks, and
all detectors exhibit different loopholes. Averaging all detectors, the
performance drops by 35% across all attacks. Further, we investigate the
reasons behind these defects and propose initial out-of-the-box patches to
improve robustness. | cs.CL | Stumbling Blocks: Stress Testing the Robustness of Machine-Generated Text Detectors Under Attacks |
2024-02-18T16:20:43Z | http://arxiv.org/pdf/2402.11633v1 | http://arxiv.org/abs/2402.11633v1 | Arian Askari, Roxana Petcu, Chuan Meng, Mohammad Aliannejadi, Amin Abolghasemi, Evangelos Kanoulas, Suzan Verberne | 2024-02-18T16:20:43Z | cs.CL | Identifying user intents in information-seeking dialogs is crucial for a
system to meet user's information needs. Intent prediction (IP) is challenging
and demands sufficient dialogs with human-labeled intents for training.
However, manually annotating intents is resource-intensive. While large
language models (LLMs) have been shown to be effective in generating synthetic
data, there is no study on using LLMs to generate intent-aware
information-seeking dialogs. In this paper, we focus on leveraging LLMs for
zero-shot generation of large-scale, open-domain, and intent-aware
information-seeking dialogs. We propose SOLID, which has novel self-seeding and
multi-intent self-instructing schemes. The former improves the generation
quality by using the LLM's own knowledge scope to initiate dialog generation;
the latter prompts the LLM to generate utterances sequentially, and mitigates
the need for manual prompt design by asking the LLM to autonomously adapt its
prompt instruction when generating complex multi-intent utterances.
Furthermore, we propose SOLID-RL, which is further trained to generate a dialog
in one step on the data generated by SOLID. We propose a length-based quality
estimation mechanism to assign varying weights to SOLID-generated dialogs based
on their quality during the training process of SOLID-RL. We use SOLID and
SOLID-RL to generate more than 300k intent-aware dialogs, surpassing the size
of existing datasets. Experiments show that IP methods trained on dialogs
generated by SOLID and SOLID-RL achieve better IP quality than ones trained on
human-generated dialogs. | cs.CL | Self-seeding and Multi-intent Self-instructing LLMs for Generating Intent-aware Information-Seeking dialogs |
2024-02-18T16:03:04Z | http://arxiv.org/pdf/2402.11628v1 | http://arxiv.org/abs/2402.11628v1 | Gleb Rodionov, Liudmila Prokhorenkova | 2024-02-18T16:03:04Z | cs.LG, cs.CL | Neural algorithmic reasoning aims to capture computations with neural
networks via learning the models to imitate the execution of classical
algorithms. While common architectures are expressive enough to contain the
correct model in the weights space, current neural reasoners are struggling to
generalize well on out-of-distribution data. On the other hand, classical
computations are not affected by distribution shifts as they can be described
as transitions between discrete computational states. In this work, we propose
to force neural reasoners to maintain the execution trajectory as a combination
of finite predefined states. Trained with supervision on the algorithm's state
transitions, such models are able to perfectly align with the original
algorithm. To show this, we evaluate our approach on the SALSA-CLRS benchmark,
where we get perfect test scores for all tasks. Moreover, the proposed
architectural choice allows us to prove the correctness of the learned
algorithms for any test data. | cs.LG | Discrete Neural Algorithmic Reasoning |
2024-02-18T15:41:31Z | http://arxiv.org/pdf/2402.11626v1 | http://arxiv.org/abs/2402.11626v1 | Yujia Zhou, Zheng Liu, Jiajie Jin, Jian-Yun Nie, Zhicheng Dou | 2024-02-18T15:41:31Z | cs.CL, cs.IR | Retrieval-augmented generation have become central in natural language
processing due to their efficacy in generating factual content. While
traditional methods employ single-time retrieval, more recent approaches have
shifted towards multi-time retrieval for multi-hop reasoning tasks. However,
these strategies are bound by predefined reasoning steps, potentially leading
to inaccuracies in response generation. This paper introduces MetaRAG, an
approach that combines the retrieval-augmented generation process with
metacognition. Drawing from cognitive psychology, metacognition allows an
entity to self-reflect and critically evaluate its cognitive processes. By
integrating this, MetaRAG enables the model to monitor, evaluate, and plan its
response strategies, enhancing its introspective reasoning abilities. Through a
three-step metacognitive regulation pipeline, the model can identify
inadequacies in initial cognitive responses and fixes them. Empirical
evaluations show that MetaRAG significantly outperforms existing methods. | cs.CL | Metacognitive Retrieval-Augmented Large Language Models |
2024-02-18T15:33:24Z | http://arxiv.org/pdf/2402.11625v1 | http://arxiv.org/abs/2402.11625v1 | Koren Lazar, Matan Vetzler, Guy Uziel, David Boaz, Esther Goldbraich, David Amid, Ateret Anaby-Tavor | 2024-02-18T15:33:24Z | cs.CL | In the digital era, the widespread use of APIs is evident. However, scalable
utilization of APIs poses a challenge due to structure divergence observed in
online API documentation. This underscores the need for automatic tools to
facilitate API consumption. A viable approach involves the conversion of
documentation into an API Specification format. While previous attempts have
been made using rule-based methods, these approaches encountered difficulties
in generalizing across diverse documentation. In this paper we introduce
SpeCrawler, a comprehensive system that utilizes large language models (LLMs)
to generate OpenAPI Specifications from diverse API documentation through a
carefully crafted pipeline. By creating a standardized format for numerous
APIs, SpeCrawler aids in streamlining integration processes within API
orchestrating systems and facilitating the incorporation of tools into LLMs.
The paper explores SpeCrawler's methodology, supported by empirical evidence
and case studies, demonstrating its efficacy through LLM capabilities. | cs.CL | SpeCrawler: Generating OpenAPI Specifications from API Documentation Using Large Language Models |
2024-02-18T15:28:39Z | http://arxiv.org/pdf/2402.11622v2 | http://arxiv.org/abs/2402.11622v2 | Junfei Wu, Qiang Liu, Ding Wang, Jinghao Zhang, Shu Wu, Liang Wang, Tieniu Tan | 2024-06-28T07:20:22Z | cs.CV, cs.AI, cs.CL, cs.LG | Object hallucination has been an Achilles' heel which hinders the broader
applications of large vision-language models (LVLMs). Object hallucination
refers to the phenomenon that the LVLMs claim non-existent objects in the
image. To mitigate the object hallucinations, instruction tuning and external
model-based detection methods have been proposed, which either require
large-scare computational resources or depend on the detection result of
external models. However, there remains an under-explored field to utilize the
LVLM itself to alleviate object hallucinations. In this work, we adopt the
intuition that the LVLM tends to respond logically consistently for existent
objects but inconsistently for hallucinated objects. Therefore, we propose a
Logical Closed Loop-based framework for Object Hallucination Detection and
Mitigation, namely LogicCheckGPT. In specific, we devise logical consistency
probing to raise questions with logical correlations, inquiring about
attributes from objects and vice versa. Whether their responses can form a
logical closed loop serves as an indicator of object hallucination. As a
plug-and-play method, it can be seamlessly applied to all existing LVLMs.
Comprehensive experiments conducted on three benchmarks across four LVLMs have
demonstrated significant improvements brought by our method, indicating its
effectiveness and generality. | cs.CV | Logical Closed Loop: Uncovering Object Hallucinations in Large Vision-Language Models |
2024-02-18T15:27:48Z | http://arxiv.org/pdf/2402.11621v3 | http://arxiv.org/abs/2402.11621v3 | Valeria Pastorino, Jasivan A. Sivakumar, Nafise Sadat Moosavi | 2024-06-15T23:20:15Z | cs.CL | Previous studies on framing have relied on manual analysis or fine-tuning
models with limited annotated datasets. However, pre-trained models, with their
diverse training backgrounds, offer a promising alternative. This paper
presents a comprehensive analysis of GPT-4, GPT-3.5 Turbo, and FLAN-T5 models
in detecting framing in news headlines. We evaluated these models in various
scenarios: zero-shot, few-shot with in-domain examples, cross-domain examples,
and settings where models explain their predictions. Our results show that
explainable predictions lead to more reliable outcomes. GPT-4 performed
exceptionally well in few-shot settings but often misinterpreted emotional
language as framing, highlighting a significant challenge. Additionally, the
results suggest that consistent predictions across multiple models could help
identify potential annotation inaccuracies in datasets. Finally, we propose a
new small dataset for real-world evaluation on headlines from a diverse set of
topics. | cs.CL | Decoding News Narratives: A Critical Analysis of Large Language Models in Framing Detection |
2024-02-18T14:57:53Z | http://arxiv.org/pdf/2402.11608v1 | http://arxiv.org/abs/2402.11608v1 | Louis Jalouzot, Robin Sobczyk, Bastien Lhopitallier, Jeanne Salle, Nur Lan, Emmanuel Chemla, Yair Lakretz | 2024-02-18T14:57:53Z | cs.CL | We introduce Metric-Learning Encoding Models (MLEMs) as a new approach to
understand how neural systems represent the theoretical features of the objects
they process. As a proof-of-concept, we apply MLEMs to neural representations
extracted from BERT, and track a wide variety of linguistic features (e.g.,
tense, subject person, clause type, clause embedding). We find that: (1)
linguistic features are ordered: they separate representations of sentences to
different degrees in different layers; (2) neural representations are organized
hierarchically: in some layers, we find clusters of representations nested
within larger clusters, following successively important linguistic features;
(3) linguistic features are disentangled in middle layers: distinct, selective
units are activated by distinct linguistic features. Methodologically, MLEMs
are superior (4) to multivariate decoding methods, being more robust to type-I
errors, and (5) to univariate encoding methods, in being able to predict both
local and distributed representations. Together, this demonstrates the utility
of Metric-Learning Encoding Methods for studying how linguistic features are
neurally encoded in language models and the advantage of MLEMs over traditional
methods. MLEMs can be extended to other domains (e.g. vision) and to other
neural systems, such as the human brain. | cs.CL | Metric-Learning Encoding Models Identify Processing Profiles of Linguistic Features in BERT's Representations |
2024-02-18T14:25:19Z | http://arxiv.org/pdf/2402.11597v2 | http://arxiv.org/abs/2402.11597v2 | Guijin Son, Sangwon Baek, Sangdae Nam, Ilgyun Jeong, Seungone Kim | 2024-06-06T12:55:17Z | cs.CL | Large language models (LLMs) are typically prompted to follow a single
instruction per inference call. In this work, we analyze whether LLMs also hold
the capability to handle multiple instructions simultaneously, denoted as
Multi-Task Inference. For this purpose, we introduce the MTI Bench(Multi-Task
Inference Benchmark), a comprehensive evaluation benchmark encompassing 5,000
instances across 25 tasks. Each task in the MTI Bench involves 2 to 3
sub-tasks. As expected, we first demonstrate that Multi-Task Inference reduces
the total inference time by 1.46 times in average since it does not require
multiple inference calls. Interestingly, contrary to the expectation that LLMs
would perform better when tasks are divided, we find that state-of-the-art
LLMs, such as Llama-2-Chat-70B and GPT-4, show up to 7.3% and 12.4% improved
performance with Multi-Task Inference compared to Single-Task Inference on the
MTI Bench. We release the MTI Bench dataset and our code at this link
https://github.com/guijinSON/MTI-Bench. | cs.CL | Multi-Task Inference: Can Large Language Models Follow Multiple Instructions at Once? |
2024-02-18T14:08:48Z | http://arxiv.org/pdf/2402.11592v3 | http://arxiv.org/abs/2402.11592v3 | Yihua Zhang, Pingzhi Li, Junyuan Hong, Jiaxiang Li, Yimeng Zhang, Wenqing Zheng, Pin-Yu Chen, Jason D. Lee, Wotao Yin, Mingyi Hong, Zhangyang Wang, Sijia Liu, Tianlong Chen | 2024-05-28T03:27:06Z | cs.LG, cs.CL | In the evolving landscape of natural language processing (NLP), fine-tuning
pre-trained Large Language Models (LLMs) with first-order (FO) optimizers like
SGD and Adam has become standard. Yet, as LLMs grow {in size}, the substantial
memory overhead from back-propagation (BP) for FO gradient computation presents
a significant challenge. Addressing this issue is crucial, especially for
applications like on-device training where memory efficiency is paramount. This
paper proposes a shift towards BP-free, zeroth-order (ZO) optimization as a
solution for reducing memory costs during LLM fine-tuning, building on the
initial concept introduced by MeZO. Unlike traditional ZO-SGD methods, our work
expands the exploration to a wider array of ZO optimization techniques, through
a comprehensive, first-of-its-kind benchmarking study across five LLM families
(Roberta, OPT, LLaMA, Vicuna, Mistral), three task complexities, and five
fine-tuning schemes. Our study unveils previously overlooked optimization
principles, highlighting the importance of task alignment, the role of the
forward gradient method, and the balance between algorithm complexity and
fine-tuning performance. We further introduce novel enhancements to ZO
optimization, including block-wise descent, hybrid training, and gradient
sparsity. Our study offers a promising direction for achieving further
memory-efficient LLM fine-tuning. Codes to reproduce all our experiments are at
https://github.com/ZO-Bench/ZO-LLM . | cs.LG | Revisiting Zeroth-Order Optimization for Memory-Efficient LLM Fine-Tuning: A Benchmark |
2024-02-18T12:50:19Z | http://arxiv.org/pdf/2402.11577v1 | http://arxiv.org/abs/2402.11577v1 | Ninglu Shao, Shitao Xiao, Zheng Liu, Peitian Zhang | 2024-02-18T12:50:19Z | cs.CL | Large language models (LLMs) call for extension of context to handle many
critical applications. However, the existing approaches are prone to expensive
costs and inferior quality of context extension. In this work, we propose
Extensible Embedding, which realizes high-quality extension of LLM's context
with strong flexibility and cost-effectiveness. Extensible embedding stand as
an enhancement of typical token embedding, which represents the information for
an extensible scope of context instead of a single token. By leveraging such
compact input units of higher information density, the LLM can access to a vast
scope of context even with a small context window. Extensible embedding is
systematically optimized in architecture and training method, which leads to
multiple advantages. 1) High flexibility of context extension, which flexibly
supports ad-hoc extension of diverse context lengths. 2) Strong sample
efficiency of training, which enables the embedding model to be learned in a
cost-effective way. 3) Superior compatibility with the existing LLMs, where the
extensible embedding can be seamlessly introduced as a plug-in component.
Comprehensive evaluations on long-context language modeling and understanding
tasks verify extensible embedding as an effective, efficient, flexible, and
compatible method to extend the LLM's context. | cs.CL | Extensible Embedding: A Flexible Multipler For LLM's Context Length |
2024-02-18T12:44:15Z | http://arxiv.org/pdf/2402.16880v2 | http://arxiv.org/abs/2402.16880v2 | Peng Xu, Wenqi Shao, Mengzhao Chen, Shitao Tang, Kaipeng Zhang, Peng Gao, Fengwei An, Yu Qiao, Ping Luo | 2024-04-19T07:54:27Z | cs.LG, cs.AI, cs.CL | Large language models (LLMs) have demonstrated outstanding performance in
various tasks, such as text summarization, text question-answering, and etc.
While their performance is impressive, the computational footprint due to their
vast number of parameters can be prohibitive. Existing solutions such as
SparseGPT and Wanda attempt to alleviate this issue through weight pruning.
However, their layer-wise approach results in significant perturbation to the
model's output and requires meticulous hyperparameter tuning, such as the
pruning rate, which can adversely affect overall model performance. To address
this, this paper introduces a novel LLM pruning technique dubbed blockwise
parameter-efficient sparsity allocation (BESA) by applying a blockwise
reconstruction loss. In contrast to the typical layer-wise pruning techniques,
BESA is characterized by two distinctive attributes: i) it targets the overall
pruning error with respect to individual transformer blocks, and ii) it
allocates layer-specific sparsity in a differentiable manner, both of which
ensure reduced performance degradation after pruning. Our experiments show that
BESA achieves state-of-the-art performance, efficiently pruning LLMs like
LLaMA1, and LLaMA2 with 7B to 70B parameters on a single A100 GPU in just five
hours. Code is available at https://github.com/OpenGVLab/LLMPrune-BESA. | cs.LG | BESA: Pruning Large Language Models with Blockwise Parameter-Efficient Sparsity Allocation |
2024-02-18T12:43:38Z | http://arxiv.org/pdf/2402.11574v1 | http://arxiv.org/abs/2402.11574v1 | Yucheng Zhou, Xiang Li, Qianning Wang, Jianbing Shen | 2024-02-18T12:43:38Z | cs.CV, cs.CL | In Large Visual Language Models (LVLMs), the efficacy of In-Context Learning
(ICL) remains limited by challenges in cross-modal interactions and
representation disparities. To overcome these challenges, we introduce a novel
Visual In-Context Learning (VICL) method comprising Visual Demonstration
Retrieval, Intent-Oriented Image Summarization, and Intent-Oriented
Demonstration Composition. Our approach retrieves images via ''Retrieval &
Rerank'' paradigm, summarises images with task intent and task-specific visual
parsing, and composes language-based demonstrations that reduce token count and
alleviate cross-modal interaction problem. Experimental evaluations on five
visual reasoning datasets demonstrate the effectiveness of our method.
Moreover, our extensive experiments leverage information flow analysis to
elucidate the effectiveness of our method, and investigate the impact of length
and position of demonstrations for LVLM. The use of in-context unlearning
further shows promise in resetting specific model knowledge without retraining. | cs.CV | Visual In-Context Learning for Large Vision-Language Models |
2024-02-18T12:41:01Z | http://arxiv.org/pdf/2402.11573v1 | http://arxiv.org/abs/2402.11573v1 | Kun Luo, Zheng Liu, Shitao Xiao, Kang Liu | 2024-02-18T12:41:01Z | cs.CL | Large language models (LLMs) call for extension of context to handle many
critical applications. However, the existing approaches are prone to expensive
costs and inferior quality of context extension. In this work, we
proposeExtensible Embedding, which realizes high-quality extension of LLM's
context with strong flexibility and cost-effectiveness. Extensible embedding
stand as an enhancement of typical token embedding, which represents the
information for an extensible scope of context instead of a single token. By
leveraging such compact input units of higher information density, the LLM can
access to a vast scope of context even with a small context window. Extensible
embedding is systematically optimized in architecture and training method,
which leads to multiple advantages. 1) High flexibility of context extension,
which flexibly supports ad-hoc extension of diverse context lengths. 2) Strong
sample efficiency of training, which enables the embedding model to be learned
in a cost-effective way. 3) Superior compatibility with the existing LLMs,
where the extensible embedding can be seamlessly introduced as a plug-in
component. Comprehensive evaluations on long-context language modeling and
understanding tasks verify extensible embedding as an effective, efficient,
flexible, and compatible method to extend the LLM's context. | cs.CL | BGE Landmark Embedding: A Chunking-Free Embedding Method For Retrieval Augmented Long-Context Large Language Models |
2024-02-18T12:36:23Z | http://arxiv.org/pdf/2402.11572v1 | http://arxiv.org/abs/2402.11572v1 | Zheng Ma, Changxin Wang, Yawen Ouyang, Fei Zhao, Jianbing Zhang, Shujian Huang, Jiajun Chen | 2024-02-18T12:36:23Z | cs.CL | Evaluating the compatibility between textual descriptions and corresponding
images represents a core endeavor within multi-modal research. In recent years,
a proliferation of reference-free methods, leveraging visual-language
pre-trained models (VLMs), has emerged. Empirical evidence has substantiated
that these innovative approaches exhibit a higher correlation with human
judgment, marking a significant advancement in the field. However, does a
higher correlation with human evaluations alone sufficiently denote the
complete of a metric? In response to this question, in this paper, we study if
there are any deficiencies in reference-free metrics. Specifically, inspired by
the Cobra Effect, we utilize metric scores as rewards to direct the captioning
model toward generating descriptions that closely align with the metric's
criteria. If a certain metric has flaws, it will be exploited by the model and
reflected in the generated sentences. Our findings reveal that descriptions
guided by these metrics contain significant flaws, e.g. incoherent statements
and excessive repetition. Subsequently, we propose a novel method termed
Self-Improving to rectify the identified shortcomings within these metrics. We
employ GPT-4V as an evaluative tool to assess generated sentences and the
result reveals that our approach achieves state-of-the-art (SOTA) performance.
In addition, we also introduce a challenging evaluation benchmark called Flaws
Caption to evaluate reference-free image captioning metrics comprehensively.
Our code is available at
https://github.com/aaronma2020/robust_captioning_metric | cs.CL | Cobra Effect in Reference-Free Image Captioning Metrics |
2024-02-18T12:35:52Z | http://arxiv.org/pdf/2402.11571v1 | http://arxiv.org/abs/2402.11571v1 | Zining Wang, Paul Reisert, Eric Nichols, Randy Gomez | 2024-02-18T12:35:52Z | cs.RO, cs.AI, cs.CL | Social robots aim to establish long-term bonds with humans through engaging
conversation. However, traditional conversational approaches, reliant on
scripted interactions, often fall short in maintaining engaging conversations.
This paper addresses this limitation by integrating large language models
(LLMs) into social robots to achieve more dynamic and expressive conversations.
We introduce a fully-automated conversation system that leverages LLMs to
generate robot responses with expressive behaviors, congruent with the robot's
personality. We incorporate robot behavior with two modalities: 1) a
text-to-speech (TTS) engine capable of various delivery styles, and 2) a
library of physical actions for the robot. We develop a custom,
state-of-the-art emotion recognition model to dynamically select the robot's
tone of voice and utilize emojis from LLM output as cues for generating robot
actions. A demo of our system is available here. To illuminate design and
implementation issues, we conduct a pilot study where volunteers chat with a
social robot using our proposed system, and we analyze their feedback,
conducting a rigorous error analysis of chat transcripts. Feedback was
overwhelmingly positive, with participants commenting on the robot's empathy,
helpfulness, naturalness, and entertainment. Most negative feedback was due to
automatic speech recognition (ASR) errors which had limited impact on
conversations. However, we observed a small class of errors, such as the LLM
repeating itself or hallucinating fictitious information and human responses,
that have the potential to derail conversations, raising important issues for
LLM application. | cs.RO | Ain't Misbehavin' -- Using LLMs to Generate Expressive Robot Behavior in Conversations with the Tabletop Robot Haru |
2024-02-18T12:33:54Z | http://arxiv.org/pdf/2402.11569v1 | http://arxiv.org/abs/2402.11569v1 | Matouš Jelínek, Eric Nichols, Randy Gomez | 2024-02-18T12:33:54Z | cs.RO, cs.AI, cs.CL | This study presents an empirical investigation into the design and impact of
autonomous dialogues in human-robot interaction for behavior change coaching.
We focus on the use of Haru, a tabletop social robot, and explore the
implementation of the Tiny Habits method for fostering positive behavior
change. The core of our study lies in developing a fully autonomous dialogue
system that maximizes Haru's emotional expressiveness and unique personality.
Our methodology involved iterative design and extensive testing of the dialogue
system, ensuring it effectively embodied the principles of the Tiny Habits
method while also incorporating strategies for trust-raising and
trust-dampening. The effectiveness of the final version of the dialogue was
evaluated in an experimental study with human participants (N=12). The results
indicated a significant improvement in perceptions of Haru's liveliness,
interactivity, and neutrality. Additionally, our study contributes to the
broader understanding of dialogue design in social robotics, offering practical
insights for future developments in the field. | cs.RO | Developing Autonomous Robot-Mediated Behavior Coaching Sessions with Haru |
2024-02-18T11:46:52Z | http://arxiv.org/pdf/2402.11550v2 | http://arxiv.org/abs/2402.11550v2 | Jun Zhao, Can Zu, Hao Xu, Yi Lu, Wei He, Yiwen Ding, Tao Gui, Qi Zhang, Xuanjing Huang | 2024-03-13T07:16:42Z | cs.CL, cs.AI | Large language models (LLMs) have demonstrated impressive performance in
understanding language and executing complex reasoning tasks. However, LLMs
with long context windows have been notorious for their expensive training
costs and high inference latency. Even the most advanced models such as GPT-4
and Claude2 often make mistakes when processing inputs of over $100k$ tokens, a
phenomenon also known as \textit{lost in the middle}. In this paper, we propose
\textsc{LongAgent}, a method based on multi-agent collaboration, which scales
LLMs (e.g., LLaMA) to a context of 128K and demonstrates potential superiority
in long-text processing compared to GPT-4. In \textsc{LongAgent}, a leader is
responsible for understanding user intent and directing team members to acquire
information from documents. Due to members' hallucinations, it is non-trivial
for a leader to obtain accurate information from the responses of dozens to
hundreds of members. To address this, we develop an \textit{inter-member
communication} mechanism to resolve response conflicts caused by hallucinations
through information sharing. Our experimental results indicate that
\textsc{LongAgent} offers a promising alternative for long-text processing. The
agent team instantiated with LLaMA-7B achieves significant improvements in
tasks such as 128k-long text retrieval, multi-hop question answering, compared
to GPT-4. | cs.CL | LongAgent: Scaling Language Models to 128k Context through Multi-Agent Collaboration |
2024-02-18T11:46:16Z | http://arxiv.org/pdf/2402.11549v2 | http://arxiv.org/abs/2402.11549v2 | Yanran Chen, Wei Zhao, Anne Breitbarth, Manuel Stoeckel, Alexander Mehler, Steffen Eger | 2024-03-28T11:16:28Z | cs.CL, cs.AI | Many studies have shown that human languages tend to optimize for lower
complexity and increased communication efficiency. Syntactic dependency
distance, which measures the linear distance between dependent words, is often
considered a key indicator of language processing difficulty and working memory
load. The current paper looks at diachronic trends in syntactic language change
in both English and German, using corpora of parliamentary debates from the
last c. 160 years. We base our observations on five dependency parsers,
including the widely used Stanford CoreNLP as well as 4 newer alternatives. Our
analysis of syntactic language change goes beyond linear dependency distance
and explores 15 metrics relevant to dependency distance minimization (DDM)
and/or based on tree graph properties, such as the tree height and degree
variance. Even though we have evidence that recent parsers trained on modern
treebanks are not heavily affected by data 'noise' such as spelling changes and
OCR errors in our historic data, we find that results of syntactic language
change are sensitive to the parsers involved, which is a caution against using
a single parser for evaluating syntactic language change as done in previous
work. We also show that syntactic language change over the time period
investigated is largely similar between English and German for the different
metrics explored: only 4% of cases we examine yield opposite conclusions
regarding upwards and downtrends of syntactic metrics across German and
English. We also show that changes in syntactic measures seem to be more
frequent at the tails of sentence length distributions. To our best knowledge,
ours is the most comprehensive analysis of syntactic language change using
modern NLP technology in recent corpora of English and German. | cs.CL | Syntactic Language Change in English and German: Metrics, Parsers, and Convergences |
2024-02-18T11:41:07Z | http://arxiv.org/pdf/2402.11548v2 | http://arxiv.org/abs/2402.11548v2 | Guijin Son, Hanwool Lee, Sungdong Kim, Seungone Kim, Niklas Muennighoff, Taekyoon Choi, Cheonbok Park, Kang Min Yoo, Stella Biderman | 2024-06-06T12:23:58Z | cs.CL | We propose KMMLU, a new Korean benchmark with 35,030 expert-level
multiple-choice questions across 45 subjects ranging from humanities to STEM.
While prior Korean benchmarks are translated from existing English benchmarks,
KMMLU is collected from original Korean exams, capturing linguistic and
cultural aspects of the Korean language. We test 27 public and proprietary LLMs
and observe the best public model to score 50.5%, leaving significant room for
improvement. This model was primarily trained for English and Chinese, not
Korean. Current LLMs tailored to Korean, such as Polyglot-Ko, perform far
worse. Surprisingly, even the most capable proprietary LLMs, e.g., GPT-4 and
HyperCLOVA X do not exceed 60%. This suggests that further work is needed to
improve LLMs for Korean, and we believe KMMLU offers the appropriate tool to
track this progress. We make our dataset publicly available on the Hugging Face
Hub and integrate the benchmark into EleutherAI's Language Model Evaluation
Harness. | cs.CL | KMMLU: Measuring Massive Multitask Language Understanding in Korean |
2024-02-18T11:24:34Z | http://arxiv.org/pdf/2402.12408v1 | http://arxiv.org/abs/2402.12408v1 | Zihao Tang, Zheqi Lv, Shengyu Zhang, Fei Wu, Kun Kuang | 2024-02-18T11:24:34Z | cs.LG, cs.AI, cs.CL | The rapid advancement of Large Language Models (LLMs) has revolutionized
various sectors by automating routine tasks, marking a step toward the
realization of Artificial General Intelligence (AGI). However, they still
struggle to accommodate the diverse and specific needs of users and simplify
the utilization of AI models for the average user. In response, we propose
ModelGPT, a novel framework designed to determine and generate AI models
specifically tailored to the data or task descriptions provided by the user,
leveraging the capabilities of LLMs. Given user requirements, ModelGPT is able
to provide tailored models at most 270x faster than the previous paradigms
(e.g. all-parameter or LoRA finetuning). Comprehensive experiments on NLP, CV,
and Tabular datasets attest to the effectiveness of our framework in making AI
models more accessible and user-friendly. Our code is available at
https://github.com/IshiKura-a/ModelGPT. | cs.LG | ModelGPT: Unleashing LLM's Capabilities for Tailored Model Generation |
2024-02-18T10:44:48Z | http://arxiv.org/pdf/2402.11542v1 | http://arxiv.org/abs/2402.11542v1 | Xinbang Dai, Huiying Li, Guilin Qi | 2024-02-18T10:44:48Z | cs.CL, cs.AI, I.2.4; I.2.7 | Spatio-temporal knowledge graphs (STKGs) extend the concept of knowledge
graphs (KGs) by incorporating time and location information. While the research
community's focus on Knowledge Graph Question Answering (KGQA), the field of
answering questions incorporating both spatio-temporal information based on
STKGs remains largely unexplored. Furthermore, a lack of comprehensive datasets
also has hindered progress in this area. To address this issue, we present
STQAD, a dataset comprising 10,000 natural language questions for
spatio-temporal knowledge graph question answering (STKGQA). Unfortunately,
various state-of-the-art KGQA approaches fall far short of achieving
satisfactory performance on our dataset. In response, we propose STCQA, a new
spatio-temporal KGQA approach that utilizes a novel STKG embedding method named
STComplEx. By extracting temporal and spatial information from a question, our
QA model can better comprehend the question and retrieve accurate answers from
the STKG. Through extensive experiments, we demonstrate the quality of our
dataset and the effectiveness of our STKGQA method. | cs.CL | Question Answering Over Spatio-Temporal Knowledge Graph |
2024-02-18T10:44:03Z | http://arxiv.org/pdf/2402.11541v3 | http://arxiv.org/abs/2402.11541v3 | Xinbang Dai, Yuncheng Hua, Tongtong Wu, Yang Sheng, Qiu Ji, Guilin Qi | 2024-06-16T14:16:56Z | cs.CL, cs.AI, I.2.4; I.2.7 | As the parameter scale of large language models (LLMs) grows, jointly
training knowledge graph (KG) embeddings with model parameters to enhance LLM
capabilities becomes increasingly costly. Consequently, the community has shown
interest in developing prompt strategies that effectively integrate KG
information into LLMs. However, the format for incorporating KGs into LLMs
lacks standardization; for instance, KGs can be transformed into linearized
triples or natural language (NL) text. Current prompting methods often rely on
a trial-and-error approach, leaving researchers with an incomplete
understanding of which KG input format best facilitates LLM comprehension of KG
content. To elucidate this, we design a series of experiments to explore LLMs'
understanding of different KG input formats within the context of prompt
engineering. Our analysis examines both literal and attention distribution
levels. Through extensive experiments, we indicate a counter-intuitive
phenomenon: when addressing fact-related questions, unordered linearized
triples are more effective for LLMs' understanding of KGs compared to fluent NL
text. Furthermore, noisy, incomplete, or marginally relevant subgraphs can
still enhance LLM performance. Finally, different LLMs have distinct
preferences for different formats of organizing unordered triples. | cs.CL | Large Language Models Can Better Understand Knowledge Graphs Than We Thought |
2024-02-18T10:36:05Z | http://arxiv.org/pdf/2402.11537v2 | http://arxiv.org/abs/2402.11537v2 | Yang Zhao, Li Du, Xiao Ding, Kai Xiong, Zhouhao Sun, Jun Shi, Ting Liu, Bing Qin | 2024-03-26T10:45:40Z | cs.CL, cs.AI | Through pretraining on a corpus with various sources, Large Language Models
(LLMs) have gained impressive performance. However, the impact of each
component of the pretraining corpus remains opaque. As a result, the
organization of the pretraining corpus is still empirical and may deviate from
the optimal. To address this issue, we systematically analyze the impact of 48
datasets from 5 major categories of pretraining data of LLMs and measure their
impacts on LLMs using benchmarks about nine major categories of model
capabilities. Our analyses provide empirical results about the contribution of
multiple corpora on the performances of LLMs, along with their joint impact
patterns, including complementary, orthogonal, and correlational relationships.
We also identify a set of ``high-impact data'' such as Books that is
significantly related to a set of model capabilities. These findings provide
insights into the organization of data to support more efficient pretraining of
LLMs. | cs.CL | Deciphering the Impact of Pretraining Data on Large Language Models through Machine Unlearning |
2024-02-18T10:28:18Z | http://arxiv.org/pdf/2403.00782v1 | http://arxiv.org/abs/2403.00782v1 | Hanshuang Tong, Jun Li, Ning Wu, Ming Gong, Dongmei Zhang, Qi Zhang | 2024-02-18T10:28:18Z | q-fin.ST, cs.AI, cs.CL | Recent advancements in large language models (LLMs) have opened new pathways
for many domains. However, the full potential of LLMs in financial investments
remains largely untapped. There are two main challenges for typical deep
learning-based methods for quantitative finance. First, they struggle to fuse
textual and numerical information flexibly for stock movement prediction.
Second, traditional methods lack clarity and interpretability, which impedes
their application in scenarios where the justification for predictions is
essential. To solve the above challenges, we propose Ploutos, a novel financial
LLM framework that consists of PloutosGen and PloutosGPT. The PloutosGen
contains multiple primary experts that can analyze different modal data, such
as text and numbers, and provide quantitative strategies from different
perspectives. Then PloutosGPT combines their insights and predictions and
generates interpretable rationales. To generate accurate and faithful
rationales, the training strategy of PloutosGPT leverage rearview-mirror
prompting mechanism to guide GPT-4 to generate rationales, and a dynamic token
weighting mechanism to finetune LLM by increasing key tokens weight. Extensive
experiments show our framework outperforms the state-of-the-art methods on both
prediction accuracy and interpretability. | q-fin.ST | Ploutos: Towards interpretable stock movement prediction with financial large language model |
2024-02-18T10:15:38Z | http://arxiv.org/pdf/2402.11534v1 | http://arxiv.org/abs/2402.11534v1 | Dayuan Fu, Jianzhao Huang, Siyuan Lu, Guanting Dong, Yejie Wang, Keqing He, Weiran Xu | 2024-02-18T10:15:38Z | cs.CL, cs.AI | Addressing the discrepancies between predictions and actual outcomes often
aids individuals in expanding their thought processes and engaging in
reflection, thereby facilitating reasoning in the correct direction. In this
paper, we introduce $\textbf{PreAct}$, an agent framework that integrates
$\textbf{pre}$diction with $\textbf{rea}$soning and $\textbf{act}$ion.
Leveraging the information provided by predictions, a large language model
(LLM) based agent can offer more diversified and strategically oriented
reasoning, which in turn leads to more effective actions that help the agent
complete complex tasks. Our experiments demonstrate that PreAct outperforms the
ReAct approach in accomplishing complex tasks and that PreAct can be
co-enhanced when combined with Reflexion methods. We prompt the model with
different numbers of historical predictions and find that historical
predictions have a sustained positive effect on LLM planning. The differences
in single-step reasoning between PreAct and ReAct show that PreAct indeed
offers advantages in terms of diversity and strategic directivity over ReAct. | cs.CL | PreAct: Predicting Future in ReAct Enhances Agent's Planning Ability |
2024-02-18T10:10:40Z | http://arxiv.org/pdf/2402.11532v2 | http://arxiv.org/abs/2402.11532v2 | Shirley Anugrah Hayati, Taehee Jung, Tristan Bodding-Long, Sudipta Kar, Abhinav Sethy, Joo-Kyung Kim, Dongyeop Kang | 2024-06-24T22:43:57Z | cs.CL | Fine-tuning large language models (LLMs) with a collection of large and
diverse instructions has improved the model's generalization to different
tasks, even for unseen tasks. However, most existing instruction datasets
include only single instructions, and they struggle to follow complex
instructions composed of multiple subtasks. In this work, we propose a novel
concept of compositional instructions called chain-of-instructions (CoI), where
the output of one instruction becomes an input for the next like a chain.
Unlike the conventional practice of solving single instruction tasks, our
proposed method encourages a model to solve each subtask step by step until the
final answer is reached. CoI-tuning (i.e., fine-tuning with CoI instructions)
improves the model's ability to handle instructions composed of multiple
subtasks as well as unseen composite tasks such as multilingual summarization.
Overall, our study find that simple CoI tuning of existing instruction data can
provide consistent generalization to solve more complex, unseen, and longer
chains of instructions. | cs.CL | Chain-of-Instructions: Compositional Instruction Tuning on Large Language Models |
2024-02-18T09:51:49Z | http://arxiv.org/pdf/2402.11525v3 | http://arxiv.org/abs/2402.11525v3 | Nuo Xu, Jun Zhao, Can Zu, Sixian Li, Lu Chen, Zhihao Zhang, Rui Zheng, Shihan Dou, Wenjuan Qin, Tao Gui, Qi Zhang, Xuanjing Huang | 2024-02-27T17:12:38Z | cs.CL, cs.LG | Faithfulness, expressiveness, and elegance is the constant pursuit in machine
translation. However, traditional metrics like \textit{BLEU} do not strictly
align with human preference of translation quality. In this paper, we explore
leveraging reinforcement learning with human feedback (\textit{RLHF}) to
improve translation quality. It is non-trivial to collect a large high-quality
dataset of human comparisons between translations, especially for low-resource
languages. To address this issue, we propose a cost-effective preference
learning strategy, optimizing reward models by distinguishing between human and
machine translations. In this manner, the reward model learns the deficiencies
of machine translation compared to human and guides subsequent improvements in
machine translation. Experimental results demonstrate that \textit{RLHF} can
effectively enhance translation quality and this improvement benefits other
translation directions not trained with \textit{RLHF}. Further analysis
indicates that the model's language capabilities play a crucial role in
preference learning. A reward model with strong language capabilities can more
sensitively learn the subtle differences in translation quality and align
better with real human translation preferences. | cs.CL | Advancing Translation Preference Modeling with RLHF: A Step Towards Cost-Effective Solution |
2024-02-18T09:42:41Z | http://arxiv.org/pdf/2402.11522v2 | http://arxiv.org/abs/2402.11522v2 | Shuai Zhang, Yu Lu, Junwen Liu, Jia Yu, Huachuan Qiu, Yuming Yan, Zhenzhong Lan | 2024-03-13T02:40:21Z | cs.CL | With the growing humanlike nature of dialog agents, people are now engaging
in extended conversations that can stretch from brief moments to substantial
periods of time. Understanding the factors that contribute to sustaining these
interactions is crucial, yet existing studies primarily focusing on short-term
simulations that rarely explore such prolonged and real conversations.
In this paper, we investigate the factors influencing retention rates in real
interactions with roleplaying models. By analyzing a large dataset of
interactions between real users and thousands of characters, we systematically
examine multiple factors and assess their impact on user retention rate.
Surprisingly, we find that the degree to which the bot embodies the roles it
plays has limited influence on retention rates, while the length of each turn
it speaks significantly affects retention rates. This study sheds light on the
critical aspects of user engagement with role-playing models and provides
valuable insights for future improvements in the development of large language
models for role-playing purposes. | cs.CL | Unveiling the Secrets of Engaging Conversations: Factors that Keep Users Hooked on Role-Playing Dialog Agents |
2024-02-18T09:21:12Z | http://arxiv.org/pdf/2402.11518v2 | http://arxiv.org/abs/2402.11518v2 | Lin Chen, Fengli Xu, Nian Li, Zhenyu Han, Meng Wang, Yong Li, Pan Hui | 2024-06-22T04:03:30Z | cs.LG, cs.CL | Heterogeneous information networks (HIN) have gained increasing popularity in
recent years for capturing complex relations between diverse types of nodes.
Meta-structures are proposed as a useful tool to identify the important
patterns in HINs, but hand-crafted meta-structures pose significant challenges
for scaling up, drawing wide research attention towards developing automatic
search algorithms. Previous efforts primarily focused on searching for
meta-structures with good empirical performance, overlooking the importance of
human comprehensibility and generalizability. To address this challenge, we
draw inspiration from the emergent reasoning abilities of large language models
(LLMs). We propose ReStruct, a meta-structure search framework that integrates
LLM reasoning into the evolutionary procedure. ReStruct uses a grammar
translator to encode the meta-structures into natural language sentences, and
leverages the reasoning power of LLMs to evaluate their semantic feasibility.
Besides, ReStruct also employs performance-oriented evolutionary operations.
These two competing forces allow ReStruct to jointly optimize the semantic
explainability and empirical performance of meta-structures. Furthermore,
ReStruct contains a differential LLM explainer to generate and refine natural
language explanations for the discovered meta-structures by reasoning through
the search history. Experiments on eight representative HIN datasets
demonstrate that ReStruct achieves state-of-the-art performance in both
recommendation and node classification tasks. Moreover, a survey study
involving 73 graduate students shows that the discovered meta-structures and
generated explanations by ReStruct are substantially more comprehensible. Our
code and questionnaire are available at https://github.com/LinChen-65/ReStruct. | cs.LG | Large Language Model-driven Meta-structure Discovery in Heterogeneous Information Network |
2024-02-18T09:10:04Z | http://arxiv.org/pdf/2402.11517v3 | http://arxiv.org/abs/2402.11517v3 | Zijin Hong, Zheng Yuan, Hao Chen, Qinggang Zhang, Feiran Huang, Xiao Huang | 2024-06-06T13:56:59Z | cs.CL | Generating accurate SQL queries for user questions (text-to-SQL) has been a
long-standing challenge since it requires a deep understanding of both the
user's question and the corresponding database schema in order to retrieve the
desired content accurately. Existing methods rely on the comprehensive
capability of large language models (LLMs) to generate the SQL. However, some
necessary knowledge is not explicitly included in the database schema and user
question or has been learned by LLMs. Thus, the generated SQL of the
knowledge-insufficient questions may be inaccurate, negatively influencing the
text-to-SQL models' performance and robustness. To address this challenge, we
propose the Knowledge-to-SQL framework, which employs tailored Data Expert LLM
(DELLM) to provide helpful knowledge for all text-to-SQL models. Specifically,
we introduce the detailed implementation of DELLM regarding table reading and
the basic fine-tuning process. We further propose a Preference Learning via
Database Feedback (PLDBF) strategy, refining the DELLM to generate more helpful
knowledge for LLMs. Extensive experiments verify that DELLM can enhance the
state-of-the-art approaches for text-to-SQL tasks. The corresponding code of
DELLM is released for further research. | cs.CL | Knowledge-to-SQL: Enhancing SQL Generation with Data Expert LLM |
2024-02-18T08:53:41Z | http://arxiv.org/pdf/2402.11512v3 | http://arxiv.org/abs/2402.11512v3 | Aishik Rakshit, Smriti Singh, Shuvam Keshari, Arijit Ghosh Chowdhury, Vinija Jain, Aman Chadha | 2024-04-16T16:40:31Z | cs.CL, cs.CY | Embeddings play a pivotal role in the efficacy of Large Language Models. They
are the bedrock on which these models grasp contextual relationships and foster
a more nuanced understanding of language and consequently perform remarkably on
a plethora of complex tasks that require a fundamental understanding of human
language. Given that these embeddings themselves often reflect or exhibit bias,
it stands to reason that these models may also inadvertently learn this bias.
In this work, we build on the seminal previous work and propose DeepSoftDebias,
an algorithm that uses a neural network to perform 'soft debiasing'. We
exhaustively evaluate this algorithm across a variety of SOTA datasets,
accuracy metrics, and challenging NLP tasks. We find that DeepSoftDebias
outperforms the current state-of-the-art methods at reducing bias across
gender, race, and religion. | cs.CL | From Prejudice to Parity: A New Approach to Debiasing Large Language Model Word Embeddings |
2024-02-18T08:32:59Z | http://arxiv.org/pdf/2402.11505v2 | http://arxiv.org/abs/2402.11505v2 | Jiamu Bai, Daoyuan Chen, Bingchen Qian, Liuyi Yao, Yaliang Li | 2024-05-30T15:46:10Z | cs.CL, cs.AI | Federated Learning (FL) has recently been applied to the parameter-efficient
fine-tuning of Large Language Models (LLMs). While promising, it raises
significant challenges due to the heterogeneous resources and data
distributions of clients. This study introduces FlexLoRA, a simple yet
effective aggregation scheme for LLM fine-tuning, which mitigates the ``bucket
effect'' in traditional FL that restricts the potential of clients with ample
resources by tying them to the capabilities of the least-resourced
participants. FlexLoRA allows for dynamic adjustment of local LoRA ranks,
fostering the development of a global model imbued with broader, less
task-specific knowledge. By synthesizing a full-size LoRA weight from
individual client contributions and employing Singular Value Decomposition
(SVD) for weight redistribution, FlexLoRA fully leverages heterogeneous client
resources. Involving thousands of clients performing heterogeneous NLP tasks
and client resources, our experiments validate the efficacy of FlexLoRA, with
the federated global model achieving consistently better improvement over SOTA
FL methods in downstream NLP task performance across various heterogeneous
distributions. FlexLoRA's practicality is further underscored by our
theoretical analysis and its seamless integration with existing LoRA-based FL
methods, offering a path toward cross-device, privacy-preserving federated
tuning for LLMs. | cs.CL | Federated Fine-tuning of Large Language Models under Heterogeneous Tasks and Client Resources |
2024-02-18T07:48:15Z | http://arxiv.org/pdf/2402.11493v2 | http://arxiv.org/abs/2402.11493v2 | Xunjian Yin, Xu Zhang, Jie Ruan, Xiaojun Wan | 2024-05-29T17:39:47Z | cs.CL | In recent years, substantial advancements have been made in the development
of large language models, achieving remarkable performance across diverse
tasks. To evaluate the knowledge ability of language models, previous studies
have proposed lots of benchmarks based on question-answering pairs. We argue
that it is not reliable and comprehensive to evaluate language models with a
fixed question or limited paraphrases as the query, since language models are
sensitive to prompt. Therefore, we introduce a novel concept named knowledge
boundary to encompass both prompt-agnostic and prompt-sensitive knowledge
within language models. Knowledge boundary avoids prompt sensitivity in
language model evaluations, rendering them more dependable and robust. To
explore the knowledge boundary for a given model, we propose projected gradient
descent method with semantic constraints, a new algorithm designed to identify
the optimal prompt for each piece of knowledge. Experiments demonstrate a
superior performance of our algorithm in computing the knowledge boundary
compared to existing methods. Furthermore, we evaluate the ability of multiple
language models in several domains with knowledge boundary. | cs.CL | Benchmarking Knowledge Boundary for Large Language Models: A Different Perspective on Model Evaluation |
2024-02-18T07:42:49Z | http://arxiv.org/pdf/2402.11489v2 | http://arxiv.org/abs/2402.11489v2 | Eran Hirsch, Guy Uziel, Ateret Anaby-Tavor | 2024-05-22T22:50:25Z | cs.CL | Planning is a fundamental task in artificial intelligence that involves
finding a sequence of actions that achieve a specified goal in a given
environment. Large language models (LLMs) are increasingly used for
applications that require planning capabilities, such as web or embodied
agents. In line with recent studies, we demonstrate through experimentation
that LLMs lack necessary skills required for planning. Based on these
observations, we advocate for the potential of a hybrid approach that combines
LLMs with classical planning methodology. Then, we introduce SimPlan, a novel
hybrid-method, and evaluate its performance in a new challenging setup. Our
extensive experiments across various planning domains demonstrate that SimPlan
significantly outperforms existing LLM-based planners. | cs.CL | What's the Plan? Evaluating and Developing Planning-Aware Techniques for Language Models |
2024-02-18T07:35:01Z | http://arxiv.org/pdf/2403.15399v1 | http://arxiv.org/abs/2403.15399v1 | Eli Bagno, Thierry Dana-Picard, Shulamit Reches | 2024-02-18T07:35:01Z | cs.CY, cs.CL, cs.LG | As soon as a new technology emerges, the education community explores its
affordances and the possibilities to apply it in education. In this paper, we
analyze sessions with ChatGPT around topics in basic Linear Algebra. We reflect
the process undertaken by the ChatGPT along the recent year in our area of
interest, emphasising the vast improvement that has been done in grappling with
Linear Algebra problems. In particular, the question whether this software can
be a teaching assistant or even somehow replace the human teacher, is
addressed. As of the time this paper is written, the answer is generally
negative. For the small part where the answer can be positive, some reflections
about an original instrumental genesis are given.
Communication with the software gives the impression to talk to a human, and
sometimes the question is whether the software understands the question or not.
Therefore, the reader's attention is drawn to the fact that ChatGPT works on a
statistical basis and not according to reflection and understanding. | cs.CY | ChatGPT in Linear Algebra: Strides Forward, Steps to Go |
2024-02-18T07:24:34Z | http://arxiv.org/pdf/2402.11485v2 | http://arxiv.org/abs/2402.11485v2 | Ikuya Yamada, Ryokan Ri | 2024-06-06T05:30:59Z | cs.CL, cs.AI, cs.LG | Adapting English-based large language models (LLMs) to other languages has
become increasingly popular due to the efficiency and potential of
cross-lingual transfer. However, existing language adaptation methods often
overlook the benefits of cross-lingual supervision. In this study, we introduce
LEIA, a language adaptation tuning method that utilizes Wikipedia entity names
aligned across languages. This method involves augmenting the target language
corpus with English entity names and training the model using left-to-right
language modeling. We assess LEIA on diverse question answering datasets using
7B-parameter LLMs, demonstrating significant performance gains across various
non-English languages. The source code is available at
https://github.com/studio-ousia/leia. | cs.CL | LEIA: Facilitating Cross-lingual Knowledge Transfer in Language Models with Entity-based Data Augmentation |
2024-02-18T07:15:03Z | http://arxiv.org/pdf/2402.14835v1 | http://arxiv.org/abs/2402.14835v1 | Jiaqi Li, Miaozeng Du, Chuanyi Zhang, Yongrui Chen, Nan Hu, Guilin Qi, Haiyun Jiang, Siyuan Cheng, Bozhong Tian | 2024-02-18T07:15:03Z | cs.CL, cs.AI, cs.LG | Multimodal knowledge editing represents a critical advancement in enhancing
the capabilities of Multimodal Large Language Models (MLLMs). Despite its
potential, current benchmarks predominantly focus on coarse-grained knowledge,
leaving the intricacies of fine-grained (FG) multimodal entity knowledge
largely unexplored. This gap presents a notable challenge, as FG entity
recognition is pivotal for the practical deployment and effectiveness of MLLMs
in diverse real-world scenarios. To bridge this gap, we introduce MIKE, a
comprehensive benchmark and dataset specifically designed for the FG multimodal
entity knowledge editing. MIKE encompasses a suite of tasks tailored to assess
different perspectives, including Vanilla Name Answering, Entity-Level Caption,
and Complex-Scenario Recognition. In addition, a new form of knowledge editing,
Multi-step Editing, is introduced to evaluate the editing efficiency. Through
our extensive evaluations, we demonstrate that the current state-of-the-art
methods face significant challenges in tackling our proposed benchmark,
underscoring the complexity of FG knowledge editing in MLLMs. Our findings
spotlight the urgent need for novel approaches in this domain, setting a clear
agenda for future research and development efforts within the community. | cs.CL | MIKE: A New Benchmark for Fine-grained Multimodal Entity Knowledge Editing |
2024-02-18T07:10:02Z | http://arxiv.org/pdf/2402.11481v1 | http://arxiv.org/abs/2402.11481v1 | YiQiu Guo, Yuchen Yang, Ya Zhang, Yu Wang, Yanfeng Wang | 2024-02-18T07:10:02Z | cs.CL | Structured data offers a sophisticated mechanism for the organization of
information. Existing methodologies for the text-serialization of structured
data in the context of large language models fail to adequately address the
heterogeneity inherent in key-value structured data. These methods are not
ideal and frequently result in larger input sizes and poor adaptability to
input changes. In this paper, we introduce DictLLM, an innovative framework
designed to improve the modeling of key-value structured data, like medical
laboratory reports, for generating medical diagnoses. DictLLM integrates three
key components: (1) group positional encoding to maintain permutation
invariance, (2) hierarchical attention bias to capture the inherent bias in
structured data, and (3) an optimal transport alignment layer that aligns the
embedding generated by the dictionary encoder with the LLM, thereby producing a
sequence of fixed-length virtual tokens. We carry out experiments using various
LLM models on a comprehensive real-world medical laboratory report dataset for
automatic diagnosis generation, our findings illustrate that DictLLM
significantly outperforms established baseline methods and few-shot GPT-4
implementations in terms of both Rouge-L and Knowledge F1 scores. Furthermore,
our evaluation of the framework's scalability and robustness, through a series
of experiments, underscores its exceptional capability in accurately modeling
the complex key-value data structure of medical dictionary data. | cs.CL | DictLLM: Harnessing Key-Value Data Structures with Large Language Models for Enhanced Medical Diagnostics |
2024-02-18T05:58:25Z | http://arxiv.org/pdf/2402.11469v2 | http://arxiv.org/abs/2402.11469v2 | Cuong Dang, Dung D. Le, Thai Le | 2024-07-02T03:29:11Z | cs.LG, cs.CL, cs.CR | Existing works have shown that fine-tuned textual transformer models achieve
state-of-the-art prediction performances but are also vulnerable to adversarial
text perturbations. Traditional adversarial evaluation is often done
\textit{only after} fine-tuning the models and ignoring the training data. In
this paper, we want to prove that there is also a strong correlation between
training data and model robustness. To this end, we extract 13 different
features representing a wide range of input fine-tuning corpora properties and
use them to predict the adversarial robustness of the fine-tuned models.
Focusing mostly on encoder-only transformer models BERT and RoBERTa with
additional results for BART, ELECTRA, and GPT2, we provide diverse evidence to
support our argument. First, empirical analyses show that (a) extracted
features can be used with a lightweight classifier such as Random Forest to
predict the attack success rate effectively, and (b) features with the most
influence on the model robustness have a clear correlation with the robustness.
Second, our framework can be used as a fast and effective additional tool for
robustness evaluation since it (a) saves 30x-193x runtime compared to the
traditional technique, (b) is transferable across models, (c) can be used under
adversarial training, and (d) robust to statistical randomness. Our code is
publicly available at \url{https://github.com/CaptainCuong/RobustText_ACL2024}. | cs.LG | A Curious Case of Searching for the Correlation between Training Data and Adversarial Robustness of Transformer Textual Models |
2024-02-18T05:40:33Z | http://arxiv.org/pdf/2402.14834v2 | http://arxiv.org/abs/2402.14834v2 | Liang Xiao, Qi Zhang, Chongyang Shi, Shoujin Wang, Usman Naseem, Liang Hu | 2024-06-19T13:15:34Z | cs.CL, cs.AI, cs.IR | The proliferation of social media platforms has fueled the rapid
dissemination of fake news, posing threats to our real-life society. Existing
methods use multimodal data or contextual information to enhance the detection
of fake news by analyzing news content and/or its social context. However,
these methods often overlook essential textual news content (articles) and
heavily rely on sequential modeling and global attention to extract semantic
information. These existing methods fail to handle the complex, subtle twists
in news articles, such as syntax-semantics mismatches and prior biases, leading
to lower performance and potential failure when modalities or social context
are missing. To bridge these significant gaps, we propose a novel multi-hop
syntax aware fake news detection (MSynFD) method, which incorporates
complementary syntax information to deal with subtle twists in fake news.
Specifically, we introduce a syntactical dependency graph and design a
multi-hop subgraph aggregation mechanism to capture multi-hop syntax. It
extends the effect of word perception, leading to effective noise filtering and
adjacent relation enhancement. Subsequently, a sequential relative
position-aware Transformer is designed to capture the sequential information,
together with an elaborate keyword debiasing module to mitigate the prior bias.
Extensive experimental results on two public benchmark datasets verify the
effectiveness and superior performance of our proposed MSynFD over
state-of-the-art detection models. | cs.CL | MSynFD: Multi-hop Syntax aware Fake News Detection |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.