Unverified50% confidenceBenchmarkExact time
On Spider, schemagate and the paper's best configuration both achieved an AC-F1 of 93.0, but schemagate's wrong-answer rate was 5.1%, slightly higher than the paper's 4.6%.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/17/2026
First Seen
Valid until: 10/31/2026
Sources
Related Entities
Related Claims
UnverifiedIn schemagate's benchmark, on the Spider dataset's 1034 questions the pure mathematical ranking accuracy is 80.3%, and 96.5% after re-ranking; on the BIRD dataset's 1534 questions the pure mathematical ranking is 88.6%, and 96.7% after re-ranking.68% similarUnverifiedIn the benchmark, Verdandi's field macro accuracy was 97.0% versus ACU's 96.2%; Verdandi's line-item F1 was 0.958 versus ACU's 0.947.64% similarUnverifiedThe Mindscend team's 26.2M-parameter GECToR-style edit annotator v1.5 achieves 0.799 on the F0.5 metric, with a 2.4% false-edit rate and a 0.0% protected-span loss rate.63% similarUnverifiedFor the schemagate benchmark's paired McNemar test, the p-value is about 1e-49 on Spider and about 1e-36 on BIRD.63% similarUnverifiedWith two to three attempts, GLM-5.3 Flash scored 82%, which is within the margin of error of GPT-5.6 Luna's 81%.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1050459API
curl https://kongchang.com/api/v1/knowledge/claims/1050459MCP
get_claim(id=1050459)