Unverified50% confidenceFactExact time
Facts located in the middle of a context are used by the model at a noticeably lower probability than those at the beginning or end — a U-shaped recall curve.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
Unverified斯坦福大学研究发现，在推理链关键步骤注入错误信息对最终答案的影响远小于预期71% similarUnverifiedThe study found that models' accuracy on semantically plausible distorted statements is actually higher than their performance on meaningless random replacements.67% similarUnverifiedWhen the context exceeds 40% of the model's context window, the model's reasoning ability noticeably degrades, exhibiting a decline in intelligence.67% similarUnverified'99% 准确率'的模型在生产环境中仍可能频繁出错，因为剩余的 1% 往往对应现实中最关键的长尾细节场景67% similarUnverifiedThe correlation between token-level probabilities (logprob) output by a model and factual accuracy is not strong, and fabricated facts may receive the same high generation probability as true information.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1002550API
curl https://kongchang.com/api/v1/knowledge/claims/1002550MCP
get_claim(id=1002550)