OneBench
Triggers and Diagnostics for LLM-Based Interpretability Failures in Active Inference Agents | OneBench: AI Insights