arXiv cs.AI· Yuxuan Chen, Mingwei Liu, Guangsheng Ou, Zekai Zhang, Zike Li, Yanlin Wang, Pelin Zheng·· 9 天前AI 评分22
MAP4CS：面向高效代码检索器微调的多维数据剪枝框架
MAP4CS: A Multi-dimensional Data Pruning Framework for Efficient Code Retriever Fine-tuning
AI 导读
MAP4CS 是一种自适应数据剪枝框架，通过融合语法结构、语义多样性与分布表示筛选高质量核心子集，并配合规则过滤流程。在两个大规模数据集上，它仅用 5% 训练数据即持续优于随机采样基线，性能可比甚至超过全量数据微调。分析显示，MAP4CS 能对冗余语料自动去重、对混乱语料去噪，构建词法多样且信息密集的训练语料。
正文 · 原文
Abstract:Retrieval-Augmented Generation (RAG) has become a cornerstone in software engineering for enhancing Large Language Models (LLMs) with domain-specific knowledge. However, adapting retrievers to evolving code repositories remains challenging due to the noise and redundancy inherent in massive code corpora. Standard fine-tuning on the full corpus is computationally expensive and often leads to sub-optimal performance due to negative transfer from low-quality samples. Conversely, simple random sampling fails to guarantee data representativeness.
To address these challenges, we propose MAP4CS (Multi-dimensional Awareness Pruning for Code Search), an adaptive data pruning framework. MAP4CS identifies a small, high-quality core subset by integrating syntactic structure, semantic diversity, and distributional representation, followed by a rigorous rule-based filtering pipeline. Extensive experiments on two large-scale datasets demonstrate that MAP4CS consistently outperforms random sampling baselines using only 5% of the training data. Remarkably, it achieves performance comparable to, or even superior to, fine-tuning on the full dataset, validating the ''less is more'' hypothesis in data-centric AI. Furthermore, linguistic analysis reveals an adaptive optimization mechanism: MAP4CS automatically functions as a de-duplicator for redundant corpora and a denoiser for chaotic ones, constructing a training corpus that is both lexically diverse and information-dense.
| Subjects: | Software Engineering (cs.SE); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.11727 [cs.SE] |
| (or arXiv:2610.11727v1 [cs.SE] for this version) | |
| https://doi.org/10.48550/arXiv.2610.11727 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yuxuan Chen [view email]
[v1]
Thu, 8 Oct 2026 11:25:50 UTC (2,692 KB)
来源：arXiv cs.AI · arxiv.org