Picture for Xiang We

Xiang We

Think Fast and Slow: Step-Level Cognitive Depth Adaptation for LLM Agents

Add code
Feb 13, 2026
Viaarxiv icon