让有害数据变有益：通过动态调整梯度方向提升模型性能
From Detrimental to Beneficial: Dynamic Influence-based Valuation and Editing
- 在批量训练中动态评估数据价值，反向调整有害样本梯度
- 显著提升分类准确率，优化过程更稳定，数据利用效率更高
- 适用于大模型微调，无需修改原始数据，兼容主流训练流程
数据估值是数据驱动学习的核心，现有方法多聚焦于识别有益或有害的训练样本，但如何利用估值结果进行后续数据干预仍研究不足。传统做法通常丢弃或降低有害样本权重，导致数据资源浪费。本文提出动态影响估值与编辑框架DIVE，可在批量层面动态估算样本价值，并将有害样本转化为有益贡献。DIVE不修改原始数据，而是在优化层面通过策略性反转有害样本的梯度方向实现干预，可无缝集成至标准学习流程且开销极小。大量实证评估表明，DIVE能持续提升分类性能，最大化数据效率，稳定优化过程，并有效推广至大语言模型微调。
原文摘要 · Abstract (English)
Data valuation is a cornerstone of data-centric learning, where prior efforts primarily focus on designing algorithms to classify training samples as either beneficial or detrimental for the learning task. However, leveraging these valuation estimates for subsequent data intervention remains underexplored; conventional approaches typically discard or downweight harmful samples, thereby underutilizing available data resources. In this paper, we present Dynamic Influence-based Valuation and Editing (DIVE), a novel and efficient framework that dynamically estimates sample values at the batch level and transforms detrimental data into beneficial contributions. Rather than altering the raw data, DIVE operates at the optimization level by strategically reversing the gradient directions of harmful samples during training, ensuring seamless integration with standard learning procedures with minimal overhead. Extensive empirical evaluations demonstrate that DIVE consistently improves classification performance, maximizes data efficiency, stabilizes optimization, and effectively generalizes to large language model fine-tuning.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。