Exact-solution programming on USACO
43.3ScoreRiVER
Evaluation Results
| Method | Links | |
|---|---|---|
| RiVERBackbone=Qwen3-8B2026.06 | 43.3 | |
| Raw-BinaryBackbone=Qwen3-8B2026.06 | 41.2 | |
| Qwen3-8BBackbone=Qwen3-8B2026.06 | 40.4 | |
| Raw-GRPOBackbone=Qwen3-8B2026.06 | 40.2 | |
| RS-GRPOBackbone=Qwen3-8B2026.06 | 40 | |
| Rank-uniformBackbone=Qwen3-8B2026.06 | 39.7 | |
| Instance-NormBackbone=Qwen3-8B2026.06 | 39.6 | |
| RiVERBackbone=GLM-Z1-9B-04142026.06 | 34.3 | |
| RS-GRPOBackbone=GLM-Z1-9B-04142026.06 | 33.2 | |
| Instance-NormBackbone=GLM-Z1-9B-04142026.06 | 33.1 | |
| Raw-GRPOBackbone=GLM-Z1-9B-04142026.06 | 32.8 | |
| Raw-BinaryBackbone=GLM-Z1-9B-04142026.06 | 32.5 | |
| GLM-Z1-9B-0414Backbone=GLM-Z1-9B-04142026.06 | 32.4 | |
| Rank-uniformBackbone=GLM-Z1-9B-04142026.06 | 31.9 |