When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning Models
当思维链更清楚时:多轮推理模型的失败模式
专题命中 推理评测 :CoT(summary_cn,abstract);reasoning(title,abstract);分类 cs.CL、cs.AI、cs.LG
AI总结 提出CoT-Output 2x2安全矩阵诊断多轮推理模型隐藏的时间动态失败,发现监督悖论和上下文注入失败两种可复现漏洞。
Comments Accepted at the ICML 2026 Workshop on Failure Modes in Agentic AI (FAGEN)