Mathematics on MATH
84.84Exact MatchN-3-Super 120B-A12B-Base
Evaluation results
| Method | Links | |
|---|---|---|
| N-3-Super 120B-A12B-BaseShots=42026.04 | 84.84 | |
| Qwen2-72B-InstructType=Instruction-tuned2024.07 | 69 | |
| Ling-flash base-2.0Shots=42026.04 | 63.8 | |
| Llama-3-70B-InstructType=Instruction-tuned2024.07 | 50.4 | |
| GLM-4.5 Air-BaseShots=42026.04 | 50.36 | |
| Mixtral-8x22B-InstructType=Instruction-tuned2024.07 | 47.4 | |
| Qwen1.5-72B-ChatType=Instruction-tuned2024.07 | 42.5 | |
| Qwen1.5-110B-ChatType=Instruction-tuned2024.07 | 42 | |
| Engram-27BShots=4-shot2026.01 | 30.7 | |
| Engram-40BShots=4-shot2026.01 | 30.6 | |
| MoE-27BShots=4-shot2026.01 | 28.3 | |
| DeepSeekMoE Chat 16B# Shot=4-shot, Total Params=16.4B, Activated Params=2.8B, FLOPs per 4K Tokens=74.4T2024.01 | 15.2 | |
| Dense-4BShots=4-shot2026.01 | 15.2 | |
| DeepSeek Chat 7B# Shot=4-shot, Total Params=6.9B, Activated Params=6.9B, FLOPs per 4K Tokens=183.5T2024.01 | 14.7 | |
| LLaMA2 SFT 7B# Shot=4-shot, Total Params=6.7B, Activated Params=6.7B, FLOPs per 4K Tokens=187.9T2024.01 | 13.5 |