When Privileged Guidance Misaligns: State-Matched Routing and Contextualized Self-Distillation for Multi-Turn Agents

• 类型：arxiv
• 标识：2608.05219
• 链接：https://arxiv.org/abs/2608.05219
• 主分类：agent
• 形态：method
• 被引：0
• 被引来源：Semantic Scholar
• S2被引：0
• 影响力被引：0
• TLDR：State-Matched Routing and Contextualized Self-Distillation (SMRC-SD), which explicitly determines when and how a privileged trajectory should guide an on-policy student, consistently outperforms unconditional successful full-path distillation.
• 待LLM分类：否
• 标题中文：特权引导失配时：面向多轮 Agent 的状态匹配路由与情境化自蒸馏
• TLDR中文：SMRC-SD（State-Matched Routing and Contextualized Self-Distillation）显式地决定特权轨迹应在何时、以何种方式指导 on-policy student，其表现始终优于无条件的成功全路径蒸馏。
• 来源文件：
• /inbox/tom/_candidates/2026-08-10-agent-rag-longcontext-candidates.json
• [S2 enrich]