| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Binary safety classification | Nemotron Response | F1 Score92 | 13 | |
| Binary safety classification | Nemotron Query | F1 Score92 | 13 | |
| English to Slovene translation | Nemotron-Chat | COMET Score0.6975 | 8 | |
| Language Modeling | Nemotron-3-Nano (val) | Validation Loss1.4122 | 6 | |
| Audio Safety Guardrail Accuracy | Nemotron Content Safety Audio | Accuracy91.3 | 6 | |
| Language Modeling | Nemotron | Perplexity14.9 | 3 | |
| Latency profiling | Nemotron-H-8B | TPOT6.8 | 3 | |
| Model Compression | Nemotron-8B | Model Size (GB)2.9 | 3 | |
| Energy consumption ranking | Nemotron 9B workload V2 | Pairwise Accuracy96.1 | 2 | |
| Tool-Risk Prediction | Nemotron core (held-out test) | Tool-Risk Accuracy (tool rows)90.3 | 2 | |
| Tool-Need Prediction | Nemotron held-out core (test) | Tool-Need Accuracy75.3 | 2 | |
| Model Ranking | Nemotron F1 (test) | Kendall Tau0.707 | 2 |