llm-explanations

研究通过预注册的受控实验比较了声明式动画规格与自由HTML/JavaScript代码两种LLM生成解释的方案，发现规格虽更短，但由于推理开销增加并未更便宜、更快或更可靠，小模型的规格可靠性甚至显著更低。

This paper formalizes the concept of explanation sufficiency for LLMs, proposes a new metric called SCSuff to evaluate free-text explanations using the model's own input beliefs, and demonstrates that current LLM explanations are generally insufficient.