List your company

Claim-level specificity control on LongFact full

11,705Claims EmittedNo CSS
7,191.48,363.29,53510,706.8Apr 19, 2026

Evaluation results

MethodLinks
No CSSPolicy=No CSS, Model=GPT-5.4, Ex.=22802026.0411,70511,70592.310092.384.6
Calibrated CSSPolicy=Calibrated, Model=GPT-5.4, Ex.=22802026.0411,08511,70598.6593.8192.5891.3
Oracle CSSPolicy=Oracle, Model=GPT-5.4, Ex.=22802026.0411,05611,70510093.5993.5993.59
Uncalibrated CSSPolicy=Uncalibrated, Model=GPT-5.4, Ex.=22802026.0410,94811,70599.3486.3385.8385.21
Claim-dropPolicy=Claim-drop, Model=GPT-5.4, Ex.=22802026.0410,82311,70598.7792.4691.3390.19
Whole abstainPolicy=Whole abstain, Model=GPT-5.4, Ex.=22802026.047,36511,70598.2562.9261.8260.72