通过筛选有效参数更新，实现高效稳定的语言模型微调
APE: Selective Fine-tuning with Acceptance Criteria for Language Model Adaptation
- 基于进化优化思想，只接受性能提升的参数修改
- 在新闻摘要任务上提升33.9% BLEU，降低36.2%困惑度
- 适合资源有限但需稳定改进的模型定制场景
我们提出相邻可能探索（APE），一种用于大语言模型适应的选择性微调方法。该方法受进化优化原理启发，通过在小数据子集上微调评估多个候选参数更新，并仅接受超过性能阈值的更新。与传统微调仅沿单一梯度方向不同，APE采用过滤选择机制，防止破坏性参数变化，同时实现系统性优化。在新闻摘要任务中，该方法实现了33.9%的BLEU提升和36.2%的困惑度降低，且计算开销极低。该方法为可控模型适配提供了实用框架，在性能提升与表征稳定性之间取得平衡。
原文摘要 · Abstract (English)
We present Adjacent Possible Exploration (APE), a selective fine-tuning method for adapting large language models that systematically explores parameter modifications while maintaining model stability. Inspired by evolutionary optimization principles, APE evaluates multiple candidate parameter updates through fine-tuning on small data subsets and accepts only those exceeding a performance threshold. Unlike standard fine-tuning that follows single gradient directions, APE implements a filtered selection process that prevents destabilizing parameter changes while enabling systematic improvement. Our method achieves 33.9\% BLEU improvement and 36.2\% perplexity reduction on news summarization tasks while using minimal computational resources. The approach provides a practical framework for controlled model adaptation that balances performance gains with representational stability.
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。