| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Code Optimization Kernels | Average Speedup8.9 | 7 | 3mo ago | ||
| HumanEval-Hard Cross-Domain | TextBFGS | Calls per Task13.6 | 5 | 3mo ago | |
| MBPP-Hard In-Domain | TextBFGS-REMO | Average Calls per Task7.1 | 5 | 3mo ago | |
| AlphaEvolve TriMul | SIA-W+H | Reward1.475 | 4 | 7d ago |