Fileinfo Classification on Custom Malware Dataset (test)
81Weighted AccuracyNemotron 3-Super 120B
Evaluation Results
| Method | Links | |
|---|---|---|
| Nemotron 3-Super 120BProtocol=Off-the-shelf, Sample size=400, Prompting=task-specific2026.06 | 81 | |
| Llama-4 Maverick-17BProtocol=Off-the-shelf, Sample size=400, Prompting=task-specific2026.06 | 77 | |
| Claude-4.5 SonnetProtocol=Off-the-shelf, Sample size=400, Prompting=task-specific2026.06 | 74 | |
| Llama-3.1-8BProtocol=PEFT, Sample size=entire test set, Fine-tuned parameters=115M2026.06 | 39 |