Dongxi 东锡 NLP
@dongxi_nlp
가입 January 2022
983 팔로잉 중 41.4K 팬
推荐，周末最佳读物！
Controlling reasoning effort in LLMs
How can an LLM switch between low-, medium-, and high-effort reasoning? And how does an LLM learn to reason more or less?
I put together a “little” article explaining how these effort levels are implemented at inference time and during training.
더 보기
