Unverified50% confidenceFactExact time
反思（Reflection）机制源于Reflexion框架，通过任务后生成自然语言经验总结实现言语强化学习
1
Sources
50%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
Related Claims
UnverifiedReflexion算法引入语言层面的自我反思机制，使Agent在每轮循环后生成自然语言评估并注入下一轮上下文73% similarUnverifiedReflexion（Shinn等，2023）引入语言强化学习概念，将任务成败信号转化为自然语言反思存入记忆缓冲区供后续参考72% similarUnverifiedReAct 让语言模型交替输出 Thought、Action、Observation 三种内容，能提升复杂任务质量并降低幻觉概率69% similarVerifiedThe effectiveness of prompt engineering is rooted in the large language model's In-Context Learning capability, which enables task switching without modifying model weights.67% similarUnverifiedBehavioral cloning is the simplest form of imitation learning, directly learning a mapping from states to actions rather than inferring the expert's reward function as inverse reinforcement learning (IRL) does.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/598497API
curl https://kongchang.com/api/v1/knowledge/claims/598497MCP
get_claim(id=598497)