arXiv subjects

Junjie Chen

Publications and source records attributed to Junjie Chen.

At least 127 records · Page 7Linked to original sources

arXiv2025

Bias Testing and Mitigation in LLM-based Code Generation

As the adoption of LLMs becomes more widespread in software coding ecosystems, a pressing issue has emerged: does the generated code contain social bias and unfairness, such as those related to age, gender, and race? This issue concerns the integrity, fairness, and ethical foundation of software applications that depend on the code generated by these models but are underexplored in the literature. This paper presents a novel bias testing framework that is specifically designed for code generation tasks. Based on this framework, we conduct an extensive empirical study on the biases in code generated by five widely studied LLMs (i.e., PALM-2-CodeChat-bison, Claude-instant-1, GPT-3.5-turbo, GPT-4-turbo, and GPT-4). Our findings reveal that biases are prevalent. For example, 13.47% to 49.10% of the codes generated by these LLMs have biased behaviors towards gender. Moreover, we study five bias mitigation prompt strategies that are commonly used in current code generation scenarios, i.e., zero-shot, one-shot, few-shot, and two Chain-of-Thought (CoT) prompts, with and without provided feedback-driven refinement. Our evaluation results illustrate that using direct prompt engineering strategies has limited effectiveness in mitigating bias, but our test execution feedback can help to reduce the ratio of code biases to a large extent (e.g., from 59.88% to 4.79% for GPT-4).

cs.SE↗

arXiv2025

Overview of the NTCIR-18 Automatic Evaluation of LLMs (AEOLLM) Task

In this paper, we provide an overview of the NTCIR-18 Automatic Evaluation of LLMs (AEOLLM) task. As large language models (LLMs) grow popular in both academia and industry, how to effectively evaluate the capacity of LLMs becomes an increasingly critical but still challenging issue. Existing methods can be divided into two types: manual evaluation, which is expensive, and automatic evaluation, which faces many limitations including task format (the majority belong to multiple-choice questions) and evaluation criteria (occupied by reference-based metrics). To advance the innovation of automatic evaluation, we propose the AEOLLM task which focuses on generative tasks and encourages reference-free methods. Besides, we set up diverse subtasks such as dialogue generation, text expansion, summary generation and non-factoid question answering to comprehensively test different methods. This year, we received 48 runs from 4 teams in total. This paper will describe the background of the task, the data set, the evaluation measures and the evaluation results, respectively.

cs.CL↗

arXiv2025

Viewport-Unaware Blind Omnidirectional Image Quality Assessment: A Flexible and Effective Paradigm

Most of existing blind omnidirectional image quality assessment (BOIQA) models rely on viewport generation by modeling user viewing behavior or transforming omnidirectional images (OIs) into varying formats; however, these methods are either computationally expensive or less scalable. To solve these issues, in this paper, we present a flexible and effective paradigm, which is viewport-unaware and can be easily adapted to 2D plane image quality assessment (2D-IQA). Specifically, the proposed BOIQA model includes an adaptive prior-equator sampling module for extracting a patch sequence from the equirectangular projection (ERP) image in a resolution-agnostic manner, a progressive deformation-unaware feature fusion module which is able to capture patch-wise quality degradation in a deformation-immune way, and a local-to-global quality aggregation module to adaptively map local perception to global quality. Extensive experiments across four OIQA databases (including uniformly distorted OIs and non-uniformly distorted OIs) demonstrate that the proposed model achieves competitive performance with low complexity against other state-of-the-art models, and we also verify its adaptive capacity to 2D-IQA.

cs.CV↗

arXiv2025

LexRAG: Benchmarking Retrieval-Augmented Generation in Multi-Turn Legal Consultation Conversation

Retrieval-augmented generation (RAG) has proven highly effective in improving large language models (LLMs) across various domains. However, there is no benchmark specifically designed to assess the effectiveness of RAG in the legal domain, which restricts progress in this area. To fill this gap, we propose LexRAG, the first benchmark to evaluate RAG systems for multi-turn legal consultations. LexRAG consists of 1,013 multi-turn dialogue samples and 17,228 candidate legal articles. Each sample is annotated by legal experts and consists of five rounds of progressive questioning. LexRAG includes two key tasks: (1) Conversational knowledge retrieval, requiring accurate retrieval of relevant legal articles based on multi-turn context. (2) Response generation, focusing on producing legally sound answers. To ensure reliable reproducibility, we develop LexiT, a legal RAG toolkit that provides a comprehensive implementation of RAG system components tailored for the legal domain. Additionally, we introduce an LLM-as-a-judge evaluation pipeline to enable detailed and effective assessment. Through experimental analysis of various LLMs and retrieval methods, we reveal the key limitations of existing RAG systems in handling legal consultation conversations. LexRAG establishes a new benchmark for the practical application of RAG systems in the legal domain, with its code and data available at https://github.com/CSHaitao/LexRAG.

cs.CL↗

arXiv2025

CaseGen: A Benchmark for Multi-Stage Legal Case Documents Generation

Legal case documents play a critical role in judicial proceedings. As the number of cases continues to rise, the reliance on manual drafting of legal case documents is facing increasing pressure and challenges. The development of large language models (LLMs) offers a promising solution for automating document generation. However, existing benchmarks fail to fully capture the complexities involved in drafting legal case documents in real-world scenarios. To address this gap, we introduce CaseGen, the benchmark for multi-stage legal case documents generation in the Chinese legal domain. CaseGen is based on 500 real case samples annotated by legal experts and covers seven essential case sections. It supports four key tasks: drafting defense statements, writing trial facts, composing legal reasoning, and generating judgment results. To the best of our knowledge, CaseGen is the first benchmark designed to evaluate LLMs in the context of legal case document generation. To ensure an accurate and comprehensive evaluation, we design the LLM-as-a-judge evaluation framework and validate its effectiveness through human annotations. We evaluate several widely used general-domain LLMs and legal-specific LLMs, highlighting their limitations in case document generation and pinpointing areas for potential improvement. This work marks a step toward a more effective framework for automating legal case documents drafting, paving the way for the reliable application of AI in the legal field. The dataset and code are publicly available at https://github.com/CSHaitao/CaseGen.

cs.CL↗

arXiv2025

Omnidirectional Image Quality Captioning: A Large-scale Database and A New Model

The fast growing application of omnidirectional images calls for effective approaches for omnidirectional image quality assessment (OIQA). Existing OIQA methods have been developed and tested on homogeneously distorted omnidirectional images, but it is hard to transfer their success directly to the heterogeneously distorted omnidirectional images. In this paper, we conduct the largest study so far on OIQA, where we establish a large-scale database called OIQ-10K containing 10,000 omnidirectional images with both homogeneous and heterogeneous distortions. A comprehensive psychophysical study is elaborated to collect human opinions for each omnidirectional image, together with the spatial distributions (within local regions or globally) of distortions, and the head and eye movements of the subjects. Furthermore, we propose a novel multitask-derived adaptive feature-tailoring OIQA model named IQCaption360, which is capable of generating a quality caption for an omnidirectional image in a manner of textual template. Extensive experiments demonstrate the effectiveness of IQCaption360, which outperforms state-of-the-art methods by a significant margin on the proposed OIQ-10K database. The OIQ-10K database and the related source codes are available at https://github.com/WenJuing/IQCaption360.

cs.CV↗

arXiv2025

Electrochemical CO2 capture with pH-independent redox chemistry

Capture of anthropogenic CO2 is critical for mitigating climate change, and reducing the energy cost is essential for wide-scale deployment. Solubility of inorganic carbon in aqueous solutions depends on the pH, and electrochemical modulation of the pH has been investigated as a means of CO2 capture and release. However, reported methods incur unavoidable energy costs due to thermodynamic penalties. In this study, we introduce a pH-independent redox chemistry that greatly lowers the thermodynamic energy costs by changing the pH without directly changing the [H+]. We show that the redox reaction of TEMPO molecules modulates the pH for capture and release of CO2 in a flow cell with an energy cost as low as 2.6 kJ/mol of CO2 corresponding to 0.027 eV/molecule. A molecular model, supported by MD and DFT simulations, is proposed of how the pH is decreased by 7.6 while largely avoiding the entropic energy cost associated with increasing the [H+]. We believe that this work showcases the potential of pH-independent redox chemistries for practical and cost-effective CO2 capture.

physics.chem-ph↗

arXiv2025

Multitask Auxiliary Network for Perceptual Quality Assessment of Non-Uniformly Distorted Omnidirectional Images

Omnidirectional image quality assessment (OIQA) has been widely investigated in the past few years and achieved much success. However, most of existing studies are dedicated to solve the uniform distortion problem in OIQA, which has a natural gap with the non-uniform distortion problem, and their ability in capturing non-uniform distortion is far from satisfactory. To narrow this gap, in this paper, we propose a multitask auxiliary network for non-uniformly distorted omnidirectional images, where the parameters are optimized by jointly training the main task and other auxiliary tasks. The proposed network mainly consists of three parts: a backbone for extracting multiscale features from the viewport sequence, a multitask feature selection module for dynamically allocating specific features to different tasks, and auxiliary sub-networks for guiding the proposed model to capture local distortion and global quality change. Extensive experiments conducted on two large-scale OIQA databases demonstrate that the proposed model outperforms other state-of-the-art OIQA metrics, and these auxiliary sub-networks contribute to improve the performance of the proposed model. The source code is available at https://github.com/RJL2000/MTAOIQA.

eess.IV↗

arXiv2024

LegalAgentBench: Evaluating LLM Agents in Legal Domain

With the increasing intelligence and autonomy of LLM agents, their potential applications in the legal domain are becoming increasingly apparent. However, existing general-domain benchmarks cannot fully capture the complexity and subtle nuances of real-world judicial cognition and decision-making. Therefore, we propose LegalAgentBench, a comprehensive benchmark specifically designed to evaluate LLM Agents in the Chinese legal domain. LegalAgentBench includes 17 corpora from real-world legal scenarios and provides 37 tools for interacting with external knowledge. We designed a scalable task construction framework and carefully annotated 300 tasks. These tasks span various types, including multi-hop reasoning and writing, and range across different difficulty levels, effectively reflecting the complexity of real-world legal scenarios. Moreover, beyond evaluating final success, LegalAgentBench incorporates keyword analysis during intermediate processes to calculate progress rates, enabling more fine-grained evaluation. We evaluated eight popular LLMs, highlighting the strengths, limitations, and potential areas for improvement of existing models and methods. LegalAgentBench sets a new benchmark for the practical application of LLMs in the legal domain, with its code and data available at \url{https://github.com/CSHaitao/LegalAgentBench}.

cs.CL↗

arXiv2024

Fixing Large Language Models' Specification Misunderstanding for Better Code Generation

Code generation is to automatically generate source code conforming to a given programming specification, which has received extensive attention especially with the development of large language models (LLMs). Due to the inherent difficulty of code generation, the code generated by LLMs may not be aligned with the specification. Although thought-eliciting prompting techniques have been proposed to enhance the code generation performance of LLMs, producing correct understanding for complicated programming problems remains challenging, resulting in unsatisfactory performance. Also, some feedback-based prompting techniques have been proposed to fix incorrect code using error messages produced by test execution. However, when the generated code deviates significantly from the ground truth, they encounter difficulties in improving performance based on such coarse-grained information. In this work, we propose a novel prompting technique, called μFiX, to improve the code generation performance of LLMs by devising both sophisticated thought-eliciting prompting and feedback-based prompting and making the first exploration on their synergy. It first exploits test case analysis to obtain specification understanding and enables a self-improvement process to identify and refine the misunderstanding in the thought-eliciting prompting phase. μFiX further fixes the specification understanding towards the direction reducing the gap between the provided understanding (from the first phase) and the actual understanding implicitly utilized by LLMs for code generation in the feedback-based prompting phase. By improving the understanding with μFiX, the code generation performance of LLMs can be largely improved. Our evaluation on two advanced LLMs (ChatGPT and DeepSeek-Coder) with six widely-used benchmarks by comparing with 15 baselines, demonstrates the effectiveness of μFiX ...

cs.SE↗

arXiv2024

ProsodyFM: Unsupervised Phrasing and Intonation Control for Intelligible Speech Synthesis

Prosody contains rich information beyond the literal meaning of words, which is crucial for the intelligibility of speech. Current models still fall short in phrasing and intonation; they not only miss or misplace breaks when synthesizing long sentences with complex structures but also produce unnatural intonation. We propose ProsodyFM, a prosody-aware text-to-speech synthesis (TTS) model with a flow-matching (FM) backbone that aims to enhance the phrasing and intonation aspects of prosody. ProsodyFM introduces two key components: a Phrase Break Encoder to capture initial phrase break locations, followed by a Duration Predictor for the flexible adjustment of break durations; and a Terminal Intonation Encoder which learns a bank of intonation shape tokens combined with a novel Pitch Processor for more robust modeling of human-perceived intonation change. ProsodyFM is trained with no explicit prosodic labels and yet can uncover a broad spectrum of break durations and intonation patterns. Experimental results demonstrate that ProsodyFM can effectively improve the phrasing and intonation aspects of prosody, thereby enhancing the overall intelligibility compared to four state-of-the-art (SOTA) models. Out-of-distribution experiments show that this prosody improvement can further bring ProsodyFM superior generalizability for unseen complex sentences and speakers. Our case study intuitively illustrates the powerful and fine-grained controllability of ProsodyFM over phrasing and intonation.

cs.CL↗

arXiv2024

LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods

The rapid advancement of Large Language Models (LLMs) has driven their expanding application across various fields. One of the most promising applications is their role as evaluators based on natural language responses, referred to as ''LLMs-as-judges''. This framework has attracted growing attention from both academia and industry due to their excellent effectiveness, ability to generalize across tasks, and interpretability in the form of natural language. This paper presents a comprehensive survey of the LLMs-as-judges paradigm from five key perspectives: Functionality, Methodology, Applications, Meta-evaluation, and Limitations. We begin by providing a systematic definition of LLMs-as-Judges and introduce their functionality (Why use LLM judges?). Then we address methodology to construct an evaluation system with LLMs (How to use LLM judges?). Additionally, we investigate the potential domains for their application (Where to use LLM judges?) and discuss methods for evaluating them in various contexts (How to evaluate LLM judges?). Finally, we provide a detailed analysis of the limitations of LLM judges and discuss potential future directions. Through a structured and comprehensive analysis, we aim aims to provide insights on the development and application of LLMs-as-judges in both research and practice. We will continue to maintain the relevant resource list at https://github.com/CSHaitao/Awesome-LLMs-as-Judges.

cs.CL↗

arXiv2024

Machine Unlearning in Forgettability Sequence

Machine unlearning (MU) is becoming a promising paradigm to achieve the "right to be forgotten", where the training trace of any chosen data points could be eliminated, while maintaining the model utility on general testing samples after unlearning. With the advancement of forgetting research, many fundamental open questions remain unanswered: do different samples exhibit varying levels of difficulty in being forgotten? Further, does the sequence in which samples are forgotten, determined by their respective difficulty levels, influence the performance of forgetting algorithms? In this paper, we identify key factor affecting unlearning difficulty and the performance of unlearning algorithms. We find that samples with higher privacy risks are more likely to be unlearning, indicating that the unlearning difficulty varies among different samples which motives a more precise unlearning mode. Built upon this insight, we propose a general unlearning framework, dubbed RSU, which consists of Ranking module and SeqUnlearn module.

cs.LG↗

arXiv2024

CalibraEval: Calibrating Prediction Distribution to Mitigate Selection Bias in LLMs-as-Judges

The use of large language models (LLMs) as automated evaluation tools to assess the quality of generated natural language, known as LLMs-as-Judges, has demonstrated promising capabilities and is rapidly gaining widespread attention. However, when applied to pairwise comparisons of candidate responses, LLM-based evaluators often exhibit selection bias. Specifically, their judgments may become inconsistent when the option positions or ID tokens are swapped, compromising the effectiveness and fairness of the evaluation result. To address this challenge, we introduce CalibraEval, a novel label-free method for mitigating selection bias during inference. Specifically, CalibraEval reformulates debiasing as an optimization task aimed at adjusting observed prediction distributions to align with unbiased prediction distributions. To solve this optimization problem, we propose a non-parametric order-preserving algorithm (NOA). This algorithm leverages the partial order relationships between model prediction distributions, thereby eliminating the need for explicit labels and precise mathematical function modeling.Empirical evaluations of LLMs in multiple representative benchmarks demonstrate that CalibraEval effectively mitigates selection bias and improves performance compared to existing debiasing methods. This work marks a step toward building more robust and unbiased automated evaluation frameworks, paving the way for improved reliability in AI-driven assessments

cs.CL↗

arXiv2024

Quantum subspace verification for error correction codes

Benchmarking the performance of quantum error correction codes in physical systems is crucial for achieving fault-tolerant quantum computing. Current methodologies, such as (shadow) tomography or direct fidelity estimation, fall short in efficiency due to the neglect of possible prior knowledge about quantum states. To address the challenge, we introduce a framework of quantum subspace verification, employing the knowledge of quantum error correction code subspaces to reduce the potential measurement budgets. Specifically, we give the sample complexity to estimate the fidelity to the target subspace under some confidence level. Building on the framework, verification operators are developed, which can be implemented with experiment-friendly local measurements for stabilizer codes and quantum low-density parity-check (QLDPC) codes. Our constructions require $O(n-k)$ local measurement settings for both, and the sample complexity of $O(n-k)$ for stabilizer codes and of $O((n-k)^2)$ for generic QLDPC codes, where $n$ and $k$ are the numbers of physical and logical qubits, respectively. Notably, for certain codes like the notable Calderbank-Shor-Steane codes and QLDPC stabilizer codes, the setting number and sample complexity can be significantly reduced and are even independent of $n$. In addition, by combining the proposed subspace verification and direct fidelity estimation, we construct a protocol to verify the fidelity of general magic logical states with exponentially smaller sample complexity than previous methods. Our finding facilitates efficient and feasible verification of quantum error correction codes and also magical states, advancing the realization in practical quantum platforms.

quant-ph↗

arXiv2024

On the Evaluation of Large Language Models in Unit Test Generation

Unit testing is an essential activity in software development for verifying the correctness of software components. However, manually writing unit tests is challenging and time-consuming. The emergence of Large Language Models (LLMs) offers a new direction for automating unit test generation. Existing research primarily focuses on closed-source LLMs (e.g., ChatGPT and CodeX) with fixed prompting strategies, leaving the capabilities of advanced open-source LLMs with various prompting settings unexplored. Particularly, open-source LLMs offer advantages in data privacy protection and have demonstrated superior performance in some tasks. Moreover, effective prompting is crucial for maximizing LLMs' capabilities. In this paper, we conduct the first empirical study to fill this gap, based on 17 Java projects, five widely-used open-source LLMs with different structures and parameter sizes, and comprehensive evaluation metrics. Our findings highlight the significant influence of various prompt factors, show the performance of open-source LLMs compared to the commercial GPT-4 and the traditional Evosuite, and identify limitations in LLM-based unit test generation. We then derive a series of implications from our study to guide future research and practical use of LLM-based unit test generation.

cs.SE↗

arXiv2024

Dependency-Aware Code Naturalness

Code naturalness, which captures repetitiveness and predictability in programming languages, has proven valuable for various code-related tasks in software engineering. However, precisely measuring code naturalness remains a fundamental challenge. Existing methods measure code naturalness over individual lines of code while ignoring the deep semantic relations among different lines, e.g., program dependency, which may negatively affect the precision of the measure. In this study, we aim to perform the first empirical study to investigate whether incorporating code dependency, instead of analyzing individual lines, can enhance the precision of measuring code naturalness. To achieve that, we first propose a new method named DAN for measuring code naturalness by incorporating the rich dependency information in the code. Specifically, DAN extracts multiple sequences of code lines by traversing the program dependency graph, where different code lines are connected by dependencies in each sequence, and then the code naturalness will be measured by taking each sequence as a whole. In this way, the dependency information can be well captured. Finally, we have conducted an extensive study to evaluate the influence of code dependency for measuring code naturalness with DAN, and compared it with the state-of-the-art methods under three emerging application scenarios of code naturalness. The results demonstrate that DAN can not only better distinguish natural and unnatural code, but also substantially boost two important downstream applications of code naturalness, i.e., distinguishing buggy and non-buggy code lines and data cleansing for training better code models, reflecting the significance of code dependency in measuring code naturalness.

cs.SE↗

arXiv2024

A Reduction from Multi-Parameter to Single-Parameter Bayesian Contract Design

The main result of this paper is an almost approximation-preserving polynomial-time reduction from the most general multi-parameter Bayesian contract design (BCD) to single-parameter BCD. That is, for any multi-parameter BCD instance $I^M$, we construct a single-parameter instance $I^S$ such that any $β$-approximate contract (resp. menu of contracts) of $I^S$ can in turn be converted to a $(β-ε)$-approximate contract (resp. menu of contracts) of $I^M$. The reduction is in time polynomial in the input size and $\log(\frac{1}ε)$; moreover, when $β= 1$ (i.e., the given single-parameter solution is exactly optimal), the dependence on $\frac{1}ε$ can be removed, leading to a polynomial-time exact reduction. This efficient reduction is somewhat surprising because in the closely related problem of Bayesian mechanism design, a polynomial-time reduction from multi-parameter to single-parameter setting is believed to not exist. Our result demonstrates the intrinsic difficulty of addressing moral hazard in Bayesian contract design, regardless of being single-parameter or multi-parameter. As byproducts, our reduction answers two open questions in recent literature of algorithmic contract design: (a) it implies that optimal contract design in single-parameter BCD is not in APX unless P=NP even when the agent's type distribution is regular, answering the open question of [Alon et al. 2021] in the negative; (b) it implies that the principal's (order-wise) tight utility gap between using a menu of contracts and a single contract is $Θ(n)$ where $n$ is the number of actions, answering the major open question of [Guruganesh et al. 2021] for the single-parameter case.

cs.GT↗