Code Generation on LiveCodeBench (Difficulty Tiers)
95.9Easy ScoreGRPO
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GRPOModel=DS 8B, Training Context=16K, Testing Context=16K2026.03 | 95.9 | 32.7 | 4.7 | 35.4 | |
| MicroCoder-GRPOModel=DS 8B, Training Context=16K, Testing Context=16K2026.03 | 95.9 | 38 | 9.7 | 39.3 | |
| MicroCoder-GRPOModel=Qwen3 4B, Training Context=4K, Testing Context=4K2026.03 | 95.3 | 35.6 | 8.1 | 37.7 | |
| DAPOModel=DS 8B, Training Context=16K, Testing Context=16K2026.03 | 95.3 | 34.6 | 6.9 | 36.9 | |
| MicroCoder-GRPOModel=Qwen3 4B, Training Context=4K, Testing Context=8K2026.03 | 93.6 | 36.1 | 10.9 | 38.7 | |
| DAPOModel=Qwen3 4B, Training Context=4K, Testing Context=8K2026.03 | 90.7 | 33.2 | 9.1 | 36.3 | |
| GRPOModel=Qwen3 4B, Training Context=4K, Testing Context=8K2026.03 | 89 | 33.7 | 9.7 | 36.3 | |
| DAPOModel=Qwen3 4B, Training Context=4K, Testing Context=4K2026.03 | 88.4 | 35.6 | 7.8 | 35.9 | |
| Qwen3-4B-InstructTesting Context=8K, RL Algorithm=None2026.03 | 87.8 | 31.7 | 8.1 | 34.7 | |
| GRPOModel=Qwen3 4B, Training Context=4K, Testing Context=4K2026.03 | 87.8 | 33.7 | 8.7 | 35.6 | |
| Qwen3-4B-InstructTesting Context=4K, RL Algorithm=None2026.03 | 86.6 | 31.7 | 7.5 | 34.1 | |
| MicroCoder-GRPOModel=Qwen3 1.7B, Training Context=4K, Testing Context=4K2026.03 | 73.3 | 14.4 | 2.2 | 23.3 | |
| MicroCoder-GRPOModel=Qwen3 1.7B, Training Context=4K, Testing Context=8K2026.03 | 70.9 | 17.8 | 2.8 | 24 | |
| DAPOModel=Qwen3 1.7B, Training Context=4K, Testing Context=4K2026.03 | 68.6 | 8.7 | 2.2 | 20.4 | |
| DAPOModel=Qwen3 1.7B, Training Context=4K, Testing Context=8K2026.03 | 68.6 | 8.7 | 2.2 | 20.4 | |
| GRPOModel=Qwen3 1.7B, Training Context=4K, Testing Context=4K2026.03 | 61 | 12.5 | 1.3 | 19.3 | |
| GRPOModel=Qwen3 1.7B, Training Context=4K, Testing Context=8K2026.03 | 61 | 12.5 | 1.3 | 19.3 | |
| Qwen3-1.7B InstructTesting Context=4K, RL Algorithm=None2026.03 | 55.2 | 7.2 | 0.9 | 16.1 | |
| Qwen3-1.7B InstructTesting Context=8K, RL Algorithm=None2026.03 | 55.2 | 7.2 | 0.9 | 16.1 |