Preprint
Aug 2026
A Graph-Based Reinforcement Learning Framework for Structured Drift Diagnosis and Recovery in Autonomous LLM Agents
This work introduces a graph-based framework in which a single small language model is trained via reinforcement learning to specialize at each node of a recovery graph, external to the main agent, external to the main agent.
Ismail El Hamraoui, Sagar Jose, Nicolas Bureau et al.
· 0 citations