ResearchBenchmarksMath Reasoning on AIME25 (Pass@16, Vendi)Follow31.7Pass@16ESamp27.5428.6229.730.78Apr 27, 2026Evaluation ResultsMethodMethodLinksPass@16Vendi ScoreESampBackbone=Qwen2.5-7B-In...Backbone=Qwen2.5-7B-Instruct, parallel sequences=162026.0431.70.46VanillaBackbone=Qwen2.5-7B-In...Backbone=Qwen2.5-7B-Instruct, parallel sequences=162026.0430.30.32Min-PBackbone=Qwen2.5-7B-In...Backbone=Qwen2.5-7B-Instruct, parallel sequences=162026.0429.50.3OverRIDEBackbone=Qwen2.5-7B-In...Backbone=Qwen2.5-7B-Instruct, parallel sequences=162026.0427.70.35