نسخة أولية وصول مفتوح
PHRBench: A Behavioral Evaluation of Post-Hallucination Reasoning in LLMs
Hallucinated information can propagate through multi-stage LLM systems and become part of the context for subsequent reasoning. Existing studies of post-hallucination reasoning (PHR) mainly characterize changes in final outcomes and aggregate reasoning dynamics, leaving how models resolve hallucinated premises at the r …