List your company

Rank Alignment on BigCodeBench Instructions-Only

1Spearman rhoConv-to-Bench
0.822160.868330.91450.96067Apr 7, 2026

Evaluation results

MethodLinks
Conv-to-BenchEvaluator=Gemini-2.5-Flash, Configuration=Instructions-Only2026.0410.00110.003
Arena-Hard-Auto (Code)Evaluator=Gemini-2.5-Flash2026.040.9430.0050.8670.017
Conv-to-BenchEvaluator=GPT-5-Mini, Configuration=Instructions-Only2026.040.8860.0190.7330.056
Arena-Hard-Auto (Code)Evaluator=GPT-5-Mini2026.040.8290.0420.7330.056