Multiple-choice reasoning on FlameBench
32.64AccuracyGLM-4
Evaluation results
| Method | Links | |
|---|---|---|
| GLM-42026.02 | 32.64 | |
| Gemini Pro2026.02 | 32.1 | |
| DeepSeek-R12026.02 | 28.37 | |
| GPT-52026.02 | 15.6 |
| Method | Links | |
|---|---|---|
| GLM-42026.02 | 32.64 | |
| Gemini Pro2026.02 | 32.1 | |
| DeepSeek-R12026.02 | 28.37 | |
| GPT-52026.02 | 15.6 |