Unverified50% confidenceBenchmarkExact time
In the experiment, unit completion accuracy increased from 72.4% to 84.9%, a gain of 12.5 percentage points.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOn the BrowseWebApp Bench benchmark, the accuracy of the fine-tuned model improved from 22% to 63%.74% similarUnverifiedIn the SST-2 dataset DistilBERT-to-DistilBERT self-distillation experiment, the reliability gating method achieved 88.800%, RAPID achieved 88.685%, representing an improvement of approximately 1.4 percentage points over the baseline of 87.271%.74% similarUnverifiedIn the BERT-to-DistilBERT distillation experiment on the AG News dataset, reliability-gated relation distillation achieved an accuracy of 94.285±0.054%, and RAPID achieved 94.241±0.025%, both outperforming the cross-entropy baseline of 94.154±0.124%74% similarUnverifiedAfter intelligent filtering, the draft acceptance rate in mathematical reasoning scenarios increased from 76.9% to 92.5%.73% similarUnverified80% of accuracy improvement comes from engineering, and 20% comes from model capability.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/898627API
curl https://kongchang.com/api/v1/knowledge/claims/898627MCP
get_claim(id=898627)