利用参数空间对称性实现大语言模型的推理技能迁移
Leveraging Parameter Space Symmetries for Reasoning Skill Transfer in LLMs
AI总结：
针对任务算术在大语言模型技能迁移中易受参数发散负干扰的问题，提出利用Transformer参数空间对称性先对齐再迁移的方法，适配GQA与SwiGLU层，可高效迁移推理技能，性能优于标准任务算术。
摘要：
任务算术是在大语言模型（LLMs）间迁移技能的有效技术，但当模型在训练过程中出现参数发散时，该技术往往会受到负干扰的影响。我们通过首先对齐模型的参数空间来解决这一局限，具体利用Transformer架构固有的置换、旋转和缩放对称性。我们针对现代的分组查询注意力（GQA）和SwiGLU层适配了参数空间对齐方法，同时探索了基于权重和基于激活的两类实现路径。采用这种先对齐的策略，我们成功将高级推理技能迁移到了不具备推理能力的模型上。在高难度推理基准上的实验表明，我们的方法表现始终优于标准任务算术。本研究为在不断迭代的大语言模型家族间融合、迁移专项技能提供了有效途径，可减少冗余微调并提升模型适配能力。
英文摘要：
Task arithmetic is a powerful technique for transferring skills between Large Language Models (LLMs), but it often suffers from negative interference when models have diverged during training. We address this limitation by first aligning the models' parameter spaces, leveraging the inherent permutation, rotation, and scaling symmetries of Transformer architectures. We adapt parameter space alignment for modern Grouped-Query Attention (GQA) and SwiGLU layers, exploring both weight-based and activation-based approaches. Using this alignment-first strategy, we successfully transfer advanced reasoning skills to a non-reasoning model. Experiments on challenging reasoning benchmarks show that our method consistently outperforms standard task arithmetic. This work provides an effective approach for merging and transferring specialized skills across evolving LLM families, reducing redundant fine-tuning and enhancing model adaptability.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
