Question Answering on Humanity's Last Exam (HLE) MCQ
19.9AccuracyMonoScale
Evaluation Results
| Method | Links | |
|---|---|---|
| MonoScaleBase Model=Qwen3-30B-A3B, Number of Agents=10, Memory Status=w/ memory2026.01 | 19.9 | |
| GPT-5Base Model=GPT-5, Number of Agents=32026.01 | 19.63 | |
| MonoScaleBase Model=Qwen3-30B-A3B, Number of Agents=7, Memory Status=w/ memory2026.01 | 18.7 | |
| MonoScaleBase Model=Qwen3-30B-A3B, Number of Agents=5, Memory Status=w/ memory2026.01 | 15.6 | |
| QWEN3-235B-A22BBase Model=QWEN3-235B-A22B, Number of Agents=32026.01 | 12.09 | |
| MonoScaleBase Model=Qwen3-30B-A3B, Number of Agents=3, Memory Status=w/o memory2026.01 | 11.6 |