Unverified50% confidenceFactExact time
多项实验表明，模型的表面推理与其实际决策机制之间存在显著偏差，通过激活修补可发现真正影响输出的内部路径与书面推理不吻合
1
Sources
50%
Confidence
Long-term
Relevance
7/19/2026
First Seen
Sources
Related Claims
VerifiedResearch has found that the chain-of-thought reasoning generated by models does not fully correspond to their actual internal computation processes, and may provide post-hoc explanations unrelated to the true decision-making path.80% similarUnverified通过激活修补（Activation Patching）等因果分析方法可以发现，真正影响模型最终输出的内部路径与书面推理并不吻合79% similarUnverified斯坦福大学等机构的研究发现，当提示中包含误导性信息时，模型往往会在思维链中合理化错误答案，而非真实记录其受到干扰的推理过程76% similarUnverifiedReasoning models have built-in autonomous decomposition and verification mechanisms, and overly detailed prompts can actually compress the model's reasoning space.76% similarUnverifiedAcademia is divided over whether chain-of-thought truly reflects a model's internal computation, with research indicating that the reasoning text output by models sometimes resembles post-hoc rationalization rather than genuine decision-making basis.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/564991API
curl https://kongchang.com/api/v1/knowledge/claims/564991MCP
get_claim(id=564991)