Commonsense Reasoning on Average Zero-shot Reasoning
61.76AccuracyQwen3-8B + CAT-Q
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-8B + CAT-Q#Tokens=1M, Quantization=W1.58A162026.06 | 61.76 | |
| TernaryLLM-8B+KD#Tokens=1T, Quantization=W1.58A162026.06 | 60.04 | |
| Qwen3-4B + CAT-Q#Tokens=2M, Quantization=W1.58A162026.06 | 58.8 | |
| TequilaLLM-3B#Tokens=10B, Quantization=W1.58A162026.06 | 57.6 | |
| BitNetV1-7B#Tokens=100B, Quantization=W1.58A162026.06 | 57.42 | |
| Qwen3-4B + CAT-Q#Tokens=1M, Quantization=W1.58A162026.06 | 57.06 | |
| TriLM-3.9B#Tokens=300B, Quantization=W1.58A162026.06 | 56.55 | |
| BitNetV1-3.9B#Tokens=100B, Quantization=W1.58A162026.06 | 54.16 | |
| Qwen3-1.7B + CAT-Q#Tokens=1M, Quantization=W1.58A162026.06 | 51.01 | |
| TriLM-1.5B#Tokens=300B, Quantization=W1.58A162026.06 | 50.2 | |
| BitNetV1-1.3B#Tokens=100B, Quantization=W1.58A162026.06 | 48.28 |