Reasoning on AIME 120
77.5AccuracyReSET
Evaluation Results
| Method | Links | |
|---|---|---|
| ReSETModel Backbone=Qwen3, Model Size=32B2026.06 | 77.5 | |
| BF16 BaselineModel Backbone=Qwen3, Model Size=14B2026.06 | 76.1 | |
| BF16 BaselineModel Backbone=Qwen3, Model Size=32B2026.06 | 75.8 | |
| 4/6Model Backbone=Qwen3, Model Size=32B2026.06 | 74.8 | |
| RTNModel Backbone=Qwen3, Model Size=32B2026.06 | 74.4 | |
| MR-GPTQModel Backbone=Qwen3, Model Size=32B2026.06 | 73.3 | |
| BRQModel Backbone=Qwen3, Model Size=32B2026.06 | 73 | |
| ReSETModel Backbone=Qwen3, Model Size=14B2026.06 | 72.1 | |
| MR-GPTQModel Backbone=Qwen3, Model Size=14B2026.06 | 71 | |
| BF16 BaselineModel Backbone=Qwen3, Model Size=8B2026.06 | 70.4 | |
| RTNModel Backbone=Qwen3, Model Size=14B2026.06 | 70.4 | |
| 4/6Model Backbone=Qwen3, Model Size=14B2026.06 | 70.1 | |
| BRQModel Backbone=Qwen3, Model Size=14B2026.06 | 66.9 | |
| MR-GPTQModel Backbone=Qwen3, Model Size=8B2026.06 | 65.2 | |
| BF16 BaselineModel Backbone=All Models, Model Size=Avg2026.06 | 65.1 | |
| ReSETModel Backbone=Qwen3, Model Size=8B2026.06 | 64.9 | |
| 4/6Model Backbone=Qwen3, Model Size=8B2026.06 | 64 | |
| RTNModel Backbone=Qwen3, Model Size=8B2026.06 | 62.5 | |
| ReSETModel Backbone=All Models, Model Size=Avg2026.06 | 62.5 | |
| 4/6Model Backbone=All Models, Model Size=Avg2026.06 | 60.6 | |
| MR-GPTQModel Backbone=All Models, Model Size=Avg2026.06 | 60 | |
| RTNModel Backbone=All Models, Model Size=Avg2026.06 | 59.9 | |
| BF16 BaselineModel Backbone=R1-Qwen, Model Size=14B2026.06 | 57.4 | |
| BRQModel Backbone=All Models, Model Size=Avg2026.06 | 57 | |
| ReSETModel Backbone=R1-Qwen, Model Size=14B2026.06 | 54 | |
| BRQModel Backbone=Qwen3, Model Size=8B2026.06 | 53.8 | |
| 4/6Model Backbone=R1-Qwen, Model Size=14B2026.06 | 53.1 | |
| RTNModel Backbone=R1-Qwen, Model Size=14B2026.06 | 52.4 | |
| MR-GPTQModel Backbone=R1-Qwen, Model Size=14B2026.06 | 50.6 | |
| BRQModel Backbone=R1-Qwen, Model Size=14B2026.06 | 49.8 | |
| BF16 BaselineModel Backbone=R1-Qwen, Model Size=7B2026.06 | 45.7 | |
| ReSETModel Backbone=R1-Qwen, Model Size=7B2026.06 | 43.8 | |
| BRQModel Backbone=R1-Qwen, Model Size=7B2026.06 | 41.4 | |
| 4/6Model Backbone=R1-Qwen, Model Size=7B2026.06 | 41.1 | |
| RTNModel Backbone=R1-Qwen, Model Size=7B2026.06 | 39.6 | |
| MR-GPTQModel Backbone=R1-Qwen, Model Size=7B2026.06 | 39.6 |