Competition Mathematical Reasoning on MATH-500 (Accuracy)
66.5AccuracyPEAR
Evaluation Results
| Method | Links | |
|---|---|---|
| PEARBackbone=Qwen-3-30B-A3B2026.05 | 66.5 | |
| ChainBackbone=Qwen-3-30B-A3B2026.05 | 62 | |
| RingBackbone=Qwen-3-30B-A3B2026.05 | 61 | |
| RandomBackbone=Qwen-3-30B-A3B2026.05 | 58.5 | |
| StarBackbone=Qwen-3-30B-A3B2026.05 | 56 | |
| CliqueBackbone=Qwen-3-30B-A3B2026.05 | 54 | |
| PEARBackbone=GPT-5.4-nano2026.05 | 52.5 | |
| PEARBackbone=Claude-Haiku-4.52026.05 | 49.5 | |
| PEARBackbone=Qwen-2.5-14B2026.05 | 48.5 | |
| RandomBackbone=GPT-5.4-nano2026.05 | 46 | |
| CliqueBackbone=GPT-5.4-nano2026.05 | 45.5 | |
| StarBackbone=GPT-5.4-nano2026.05 | 44.5 | |
| CoT-SCBackbone=Qwen-3-30B-A3B2026.05 | 44 | |
| RingBackbone=GPT-5.4-nano2026.05 | 44 | |
| ChainBackbone=GPT-5.4-nano2026.05 | 43.5 | |
| CliqueBackbone=Claude-Haiku-4.52026.05 | 42.5 | |
| ChainBackbone=Qwen-2.5-14B2026.05 | 42 | |
| RandomBackbone=Claude-Haiku-4.52026.05 | 42 | |
| RingBackbone=Claude-Haiku-4.52026.05 | 41.5 | |
| RingBackbone=Qwen-2.5-14B2026.05 | 41 | |
| StarBackbone=Claude-Haiku-4.52026.05 | 41 | |
| RandomBackbone=Qwen-2.5-14B2026.05 | 40.5 | |
| ChainBackbone=Claude-Haiku-4.52026.05 | 40.5 | |
| StarBackbone=Qwen-2.5-14B2026.05 | 39.5 | |
| CliqueBackbone=Qwen-2.5-14B2026.05 | 38.5 | |
| CoT-SCBackbone=Claude-Haiku-4.52026.05 | 35.5 | |
| CoTBackbone=Qwen-3-30B-A3B2026.05 | 34.5 | |
| CoT-SCBackbone=GPT-5.4-nano2026.05 | 34 | |
| PEARBackbone=Gemma-3-12B2026.05 | 33.5 | |
| CoTBackbone=GPT-5.4-nano2026.05 | 31.5 | |
| RandomBackbone=Gemma-3-12B2026.05 | 30 | |
| ChainBackbone=Gemma-3-12B2026.05 | 29.5 | |
| RingBackbone=Gemma-3-12B2026.05 | 29 | |
| StarBackbone=Gemma-3-12B2026.05 | 28.5 | |
| CoTBackbone=Claude-Haiku-4.52026.05 | 28.5 | |
| CliqueBackbone=Gemma-3-12B2026.05 | 27.5 | |
| CoT-SCBackbone=Qwen-2.5-14B2026.05 | 27.5 | |
| PEARBackbone=Llama-3.1-8B2026.05 | 25.5 | |
| CoTBackbone=Qwen-2.5-14B2026.05 | 25.5 | |
| ChainBackbone=Llama-3.1-8B2026.05 | 22.5 | |
| RandomBackbone=Llama-3.1-8B2026.05 | 22 | |
| RingBackbone=Llama-3.1-8B2026.05 | 21.5 | |
| CoT-SCBackbone=Gemma-3-12B2026.05 | 20.5 | |
| CliqueBackbone=Llama-3.1-8B2026.05 | 20.5 | |
| StarBackbone=Llama-3.1-8B2026.05 | 19.5 | |
| CoTBackbone=Gemma-3-12B2026.05 | 18.5 | |
| CoT-SCBackbone=Llama-3.1-8B2026.05 | 16.5 | |
| CoTBackbone=Llama-3.1-8B2026.05 | 13.5 |