PulseAugur
中
实时 06:44:45

New Anchored Learning framework stabilizes LLM fine-tuning, cuts catastrophic forgetting

作者 PulseAugur 编辑部 · [2 个来源] ·

Researchers have developed a new framework called Anchored Learning to mitigate catastrophic forgetting in large language models during supervised fine-tuning. This method explicitly controls distributional updates by using a dynamic moving anchor, which interpolates between the current and a frozen reference model. The approach theoretically guarantees stable transitions between model distributions and empirically demonstrates significant reductions in performance degradation on benchmarks like iGSM and MedCalc, while maintaining near-optimal gains. AI

影响 Addresses catastrophic forgetting in LLMs, potentially improving the stability and reliability of fine-tuned models.

排序理由 The cluster contains an arXiv preprint detailing a new method for stabilizing LLM fine-tuning.

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

New Anchored Learning framework stabilizes LLM fine-tuning, cuts catastrophic forgetting

报道来源 [2]

  1. arXiv cs.LG TIER_1 English(EN) · Xinyu Wang, Changzhi Sun, Yuanbin Wu, Xiaoling Wang ·

    Stabilizing LLM Supervised Fine-Tuning via Explicit Distributional Control

    arXiv:2605.04468v1 Announce Type: new Abstract: Post-training large language models (LLMs) often suffers from catastrophic forgetting, where improvements on a target objective degrade previously acquired capabilities. Recent evidence suggests that this phenomenon is primarily dri…

  2. arXiv cs.CL TIER_1 English(EN) · Xiaoling Wang ·

    Stabilizing LLM Supervised Fine-Tuning via Explicit Distributional Control

    Post-training large language models (LLMs) often suffers from catastrophic forgetting, where improvements on a target objective degrade previously acquired capabilities. Recent evidence suggests that this phenomenon is primarily driven by excessive distributional drift during opt…

相关实体

相关话题