Unverified50% confidenceFactExact time
In-distribution means the test data and training data come from the same distribution; out-of-distribution means the test data come from a different distribution (such as harder, longer, or structurally different tasks).
1
Sources
50%
Confidence
Long-term
Relevance
11/3/2026
First Seen
Sources
Related Entities
Related Claims
Unverified参数量相似的情况下，训练数据质量分布的差异往往比架构差异更能决定特定场景的表现上限76% similarUnverified数据增强的核心原则是增强分布应当匹配真实测试分布，理想策略是让训练变换分布成为真实测试分布的一个超集75% similarUnverifiedWhen saving checkpoints during distributed training, the data loader state and random seeds must also be saved; otherwise, after recovery the data order and augmentation results will be inconsistent, making training irreproducible.71% similarUnverifiedBy projecting input activations onto a small subset of the principal components of the training distribution, probes can achieve cross-domain transfer with performance nearly matching that of probes trained directly on the test distribution.70% similarUnverifiedEvaluating within the training distribution easily overestimates model capability, while out-of-distribution benchmarks require the model to generalize to unseen scene and interaction combinations, better exposing the gap between 'looking right' and 'being physically right'.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1130339API
curl https://kongchang.com/api/v1/knowledge/claims/1130339MCP
get_claim(id=1130339)