Unverified50% confidenceBenchmarkExact time
One SJR training variant uses no real violating samples at all (all positive-class data is synthetically generated), and its violation-class F1 score differs from the model trained on full real data by less than 0.2% in relative terms.
1
Sources
50%
Confidence
Long-term
Relevance
9/22/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOn a misleading advertisement detection task, SJR achieved a 23.6% relative improvement in F1 score on the 'non-misleading' category compared to the zero-shot chain-of-thought baseline.67% similarUnverified即使只有5-10%的训练样本存在错位，也会导致VLM模型在下游任务上出现显著的性能退化64% similarUnverifiedWhen actually training Transformer models, the Model FLOPS Utilization (MFU) is typically only 30%–60%.63% similarUnverifiedExperiments conducted by the research team on models such as GPT-2 showed that after only 5–9 generations of recursive training, the quality of model outputs degrades significantly.63% similarUnverifiedFor cell types or species underrepresented in the training data, the reliability of predictions from such models decreases significantly.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/940019API
curl https://kongchang.com/api/v1/knowledge/claims/940019MCP
get_claim(id=940019)