Test-Time Learning for Large Language Models

Jinwu Hu, Zhitian Zhang, Guohao Chen +6

While Large Language Models (LLMs) have exhibited remarkable emergent capabilities through extensive pre-training, they still face critical limitations in generalizing to specializ…