AI RESEARCH
Verify Before You Commit: Towards Faithful Reasoning in LLM Agents via Self-Auditing
arXiv CS.CL
•
ArXi:2604.08401v1 Announce Type: cross In large language model (LLM) agents, reasoning trajectories are treated as reliable internal beliefs for guiding actions and updating memory. However, coherent reasoning can still violate logical or evidential constraints, allowing uned beliefs repeatedly d and propagated across decision steps, leading to systematic behavioral drift in long-horizon agentic systems. Most existing strategies rely on the consensus mechanism, conflating agreement with faithfulness.