Beliefs replace full interaction history as working context
Belief states replace the full interaction history as the agent’s working context; belief grading supervises the content of each belief state to improve performance.
Evidencia a favor
Teaching LLMs to Update Beliefs for Efficient Long-Horizon Interaction
Extracto original
Beliefs replace the full interaction history as the agent’s working context, and belief grading improves performance by supervising the contents of each belief state.
Contexto
Overview of ABBEL compared to traditional recursive summarization. .