List your company

Code Generation on Hard Corpus 192 problems 4

83.3Correctness (pass@1)AlgoSkill
51.37259.66167.9576.239Jun 29, 2026

Evaluation results

MethodLinks
AlgoSkillBackbone=GPT-5.52026.0683.3
DirectBackbone=GPT-5.52026.0681.8
CoTBackbone=Sonnet-4-52026.0677.1
AlgoSkillBackbone=Sonnet-4-52026.0677.1
CoTBackbone=GPT-5.52026.0676.6
DirectBackbone=Sonnet-4-52026.0675
AlgoSkillBackbone=gpt-oss-120b2026.0675
AlgoSkillBackbone=Haiku2026.0671.4
DirectBackbone=Haiku2026.0670.3
CoTBackbone=Haiku2026.0667.7
DirectBackbone=gpt-oss-120b2026.0654.7
CoTBackbone=gpt-oss-120b2026.0652.6