pith. sign in

arxiv: 2502.06215 · v1 · pith:LKYQ57X3new · submitted 2025-02-10 · 💻 cs.SE · cs.AI· cs.CL

LessLeak-Bench: A First Investigation of Data Leakage in LLMs Across 83 Software Engineering Benchmarks

classification 💻 cs.SE cs.AIcs.CL
keywords leakagedatabenchmarksllmsbenchmarkevaluationratiosresearch
0
0 comments X
read the original abstract

Large Language Models (LLMs) are widely utilized in software engineering (SE) tasks, such as code generation and automated program repair. However, their reliance on extensive and often undisclosed pre-training datasets raises significant concerns about data leakage, where the evaluation benchmark data is unintentionally ``seen'' by LLMs during the model's construction phase. The data leakage issue could largely undermine the validity of LLM-based research and evaluations. Despite the increasing use of LLMs in the SE community, there is no comprehensive study that assesses the extent of data leakage in SE benchmarks for LLMs yet. To address this gap, this paper presents the first large-scale analysis of data leakage in 83 SE benchmarks concerning LLMs. Our results show that in general, data leakage in SE benchmarks is minimal, with average leakage ratios of only 4.8\%, 2.8\%, and 0.7\% for Python, Java, and C/C++ benchmarks, respectively. However, some benchmarks exhibit relatively higher leakage ratios, which raises concerns about their bias in evaluation. For instance, QuixBugs and BigCloneBench have leakage ratios of 100.0\% and 55.7\%, respectively. Furthermore, we observe that data leakage has a substantial impact on LLM evaluation. We also identify key causes of high data leakage, such as the direct inclusion of benchmark data in pre-training datasets and the use of coding platforms like LeetCode for benchmark construction. To address the data leakage, we introduce \textbf{LessLeak-Bench}, a new benchmark that removes leaked samples from the 83 SE benchmarks, enabling more reliable LLM evaluations in future research. Our study enhances the understanding of data leakage in SE benchmarks and provides valuable insights for future research involving LLMs in SE.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. The Rise of AI Teammates in Software Engineering (SE) 3.0: How Autonomous Coding Agents Are Reshaping Software Engineering

    cs.SE 2025-07 conditional novelty 8.0

    AIDev is a new open dataset of 456k AI-agent pull requests showing agents submit code faster than humans but with lower acceptance rates and simpler changes.

  2. Guidelines for Empirical Studies in Software Engineering involving Large Language Models

    cs.SE 2025-08 accept novelty 7.0

    The paper delivers a taxonomy of seven LLM study types in software engineering along with eight guidelines that separate mandatory requirements from recommended practices to address reproducibility challenges.

  3. BT-APE: A Computationally Light Backtracking Approach to Automatic Prompt Engineering for Requirements Classification

    cs.SE 2026-07 unverdicted novelty 6.0

    BT-APE automates prompt engineering for requirements classification using backtracking search and dynamic examples, matching PE2 accuracy while using 72% fewer tokens and 66% less time than that baseline.

  4. SrDetection: A Self-Referential Framework for Data Leakage Detection in Code Large Language Models

    cs.CL 2026-06 unverdicted novelty 6.0

    SrDetection detects data leakage in Code LLMs via contrast between original benchmark samples and their semantic variants, reporting F1 gains of 21.52 (gray-box) and 14.46 (black-box) over baselines in a controlled testbed.

  5. Benchmarking Empirical Privacy Protection for Adaptations of Large Language Models

    cs.LG 2026-06 unverdicted novelty 6.0

    Empirical benchmarks show distribution similarity between adaptation and pretraining data increases practical privacy leakage in DP-adapted LLMs at fixed theoretical guarantees, with LoRA providing strongest protectio...

  6. PRISM: Generation-Time Detection and Mitigation of Secret Leakage in Multi-Agent LLM Pipelines

    cs.AI 2026-05 unverdicted novelty 6.0

    PRISM detects and stops credential leakage during LLM generation in multi-agent pipelines using per-token risk scores from lexical, structural, and behavioral signals, achieving zero observed leaks and F1 of 0.832 on ...

  7. Guidelines for Empirical Studies in Software Engineering involving Large Language Models

    cs.SE 2025-08 accept novelty 6.0

    A group of 22 researchers proposes seven study types and eight guidelines for empirical software engineering studies involving LLMs to enhance reproducibility and replicability.

  8. OpenAnt: LLM-Powered Vulnerability Discovery Through Code Decomposition, Adversarial Verification, and Dynamic Testing

    cs.CR 2026-06 unverdicted novelty 5.0

    OpenAnt is an open-source pipeline that uses code decomposition, LLM-based adversarial verification, and automated dynamic testing to find vulnerabilities in large projects like OpenSSL and WordPress while claiming lo...

  9. Linguistic Productivity in Large Language Models: Models Coerce, but do not Preempt

    cs.CL 2026-06 unverdicted novelty 5.0

    Larger LLMs reproduce constructional productivity via entrenchment in coercion cases with nonce words but fail to use statistical preemption to avoid overgeneralizing semantically plausible but unobserved patterns.