Unverified50% confidenceSolutionExact time
Out-of-sample design can substantially eliminate the interference of memorization retrieval on evaluation by using contemporary works created after the model's training cutoff date or from controlled sources.
1
Sources
50%
Confidence
Long-term
Relevance
9/18/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedBenchmark data contamination refers to the phenomenon where, after a test set is made public, its questions may appear in the pre-training corpora of subsequent models, causing scores to reflect memorization rather than genuine reasoning ability.75% similarUnverifiedDuplicate content in training data can cause model memorization and performance degradation.73% similarUnverifiedThe closer an event is to the cutoff date, the smaller its share in the training corpus, and the less clearly the model has mastered it.73% similarUnverified差分隐私训练、去重预处理和机器遗忘等缓解记忆化的技术方案均会在不同程度上牺牲模型性能72% similarUnverifiedThe PlanFence executor validates the records referenced by a plan before executing external actions, replans once if the records have been updated, and blocks if validation cannot be completed.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/931197API
curl https://kongchang.com/api/v1/knowledge/claims/931197MCP
get_claim(id=931197)