arXivLLM
Minimally Invasive Steering of LMs: Optimizing Rewards Without Quality Degradation
微創型語言模型導向技術:利用 MISVO 在不損害生成品質下優化輸出
This paper introduces MISVO, a minimally invasive steering method that uses local KL geometry to optimize LLM outputs for test-time rewards without parameter updates or quality degradation.
2 min read