List your company

Competitive Programming on Hard Bench HB-15 v2

47CorrectnessReflexion
26.231.63742.4Jun 29, 2026

Evaluation results

MethodLinks
ReflexionBackbone=Haiku-4-5, Evaluation Protocol=pass@52026.064786
AlgoSkill (MCTS)Backbone=Haiku-4-5, Evaluation Protocol=pass@52026.0647100
Self-RefineBackbone=Haiku-4-5, Evaluation Protocol=pass@52026.0627—