Programming Reasoning on LiveCodeBench v5
66pass@1Klear-Reasoner-8B
Evaluation Results
| Method | Links | |
|---|---|---|
| Klear-Reasoner-8BMaximum Inference Length=64K2025.08 | 66 | |
| OpenReasoning-Nemotron-7BMaximum Inference Length=64K2025.08 | 65.6 | |
| InfLLM-v2Model Size=14B, Temperature=0.6, Max output length=32k, Training context length=32k, Training tokens=1.2T, Sampling strategy=pass@12026.01 | 64.2 | |
| SPLAModel Size=14B, Temperature=0.6, Max output length=32k, Training context length=32k, Training tokens=1.2T, Sampling strategy=pass@12026.01 | 62.4 | |
| SPAModel Size=14B, Temperature=0.6, Max output length=32k, Training context length=32k, Training tokens=1.2T, Sampling strategy=pass@12026.01 | 62 | |
| Dense AttentionModel Size=14B, Temperature=0.6, Max output length=32k, Training context length=32k, Training tokens=1.2T, Sampling strategy=pass@12026.01 | 61.6 | |
| Klear-Reasoner-8BMaximum Inference Length=32K2025.08 | 61.6 | |
| Deepseek-R1-0528-Distill-8BMaximum Inference Length=64K2025.08 | 61 | |
| POLARIS-4B-PreviewMaximum Inference Length=96K2025.08 | 58.5 | |
| Klear-Reasoner-8B-SFTMaximum Inference Length=32K2025.08 | 58.5 | |
| MiMo-7B-RLMaximum Inference Length=32K2025.08 | 57.8 | |
| Qwen3-8BMaximum Inference Length=32K2025.08 | 57.5 | |
| AceReason-Nemotron-1.1-7BMaximum Inference Length=32K2025.08 | 57.2 | |
| NSAModel Size=14B, Temperature=0.6, Max output length=32k, Training context length=32k, Training tokens=1.2T, Sampling strategy=pass@12026.01 | 49.1 | |
| Skywork-OR1-7BMaximum Inference Length=32K2025.08 | 47.6 | |
| AReal-boba-RL-7BMaximum Inference Length=32K2025.08 | 34.3 |