Program Synthesis on HumanEval Standard Relaxed (test)
0.988pass@1QualityFlow
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| QualityFlowLLM Backbone=Claude Sonnet-3.5-v2, Evaluation Setting=Relaxed2025.01 | 0.988 | 0.6 | |
| LDB2025.01 | 0.982 | — | |
| QualityFlowLLM Backbone=Claude Sonnet-3.5-v2, Evaluation Setting=Standard2025.01 | 0.976 | — |