Graduate-level Science Reasoning on GPQA-D (Accuracy)
66.5AccuracyBF16
Evaluation results
| Method | Links | |
|---|---|---|
| BF16Model=Llama Nemotron Super V12026.01 | 66.5 | |
| NVFP4 QADModel=Llama Nemotron Super V12026.01 | 64.5 | |
| BF16Model=Nemotron Nano V22026.01 | 64 | |
| NVFP4 QATModel=Llama Nemotron Super V12026.01 | 63.3 | |
| NVFP4 QADModel=Nemotron Nano V22026.01 | 62.7 | |
| NVFP4 PTQModel=Llama Nemotron Super V12026.01 | 62.1 | |
| NVFP4 PTQModel=Nemotron Nano V22026.01 | 59 | |
| NVFP4 QATModel=Nemotron Nano V22026.01 | 56.9 |