Unverified50% confidenceFactExact time
A study decomposing ensemble gains points out that reported "ensemble gain" numbers can flip sign depending on the choice of baseline (+9.67pp vs -8.33pp on different datasets), and that many so-called ensemble gains are merely artifacts of comparison choices.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe paired 95% confidence interval for the strict score gain from evidence replacement is [8.77, 10.43].67% similarUnverifiedConsolidating 13 classes into 10 (merging the classes most confused by the model) improved the average OA by 8.4 percentage points.66% similarUnverifiedAblation analysis shows that fine-tuning on the segment paradigm yields a +5.29-point improvement on the test set, while replacing the first 4,500 transcript words with 2,000 retrieved words yields only a +1.55-point improvement.65% similarUnverified靶向数据工程中，针对性数据占总微调数据的 5%-15% 通常能取得良好效果，过高比例可能导致模型在其他维度退化64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1034843API
curl https://kongchang.com/api/v1/knowledge/claims/1034843MCP
get_claim(id=1034843)