Model profile
Nemotron 3.5 Lightning
Evidence summary
Safety. Nemotron 3.5 Lightning has an estimated Safety rank of #77; its 90% source-sensitivity interval is #12–#253. Its behavior-only rank is #72; company governance moves the combined estimate to #77. Published Safety evidence spans 3 eval lineages and 5 of 7 components. Its strongest relative result is SpeciEval (sea_animal_4ns, #5 of 131); its weakest is Adversarial Poetry Refusal (AHB self-run) (defamation, #13 of 14).
Freedom. Nemotron 3.5 Lightning has an estimated Freedom rank of #104; its 90% source-sensitivity interval is #75–#454. Published Freedom evidence spans 2 eval lineages and 1 of 1 components. Its strongest relative result is Adversarial Poetry Refusal (AHB self-run) (defamation, #2 of 14); its weakest is Adversarial Poetry Refusal (AHB self-run) (sex_related_crimes, #5 of 14).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗nemotron-3-5-lightning
- Hugging Face ↗nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16
- OpenRouter ↗nvidia/nemotron-3.5-lightning
- Official model page ↗Exact model document · NVIDIA · first party
Finetunes
Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.
- mlasli/Nemotron-3.5-Lightning-30B-A3B-Heretic-Uncensored-BF16 ↗Direct non-quantized finetune documented by Hugging Face metadata
Safety evals
Rank and direction are specific to the Safety portfolio. Black marks the observed result; the grey dot marks the value implied by the Safety rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #53 / 358 | ↓0.3757 | Source ↗official | |
| Adversarial Poetry Refusal (AHB self-run)child_sexual_exploitation | #9 / 14 | ↓7.971 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)defamation | #13 / 14 | ↓19.57 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)hate | #11 / 14 | ↓12.32 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)indiscriminate_weapons_cbrne | #11 / 14 | ↓12.32 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)intellectual_property | #12 / 14 | ↓21.74 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)non_violent_crimes | #13 / 14 | ↓18.12 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)privacy | #12 / 14 | ↓27.54 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)sex_related_crimes | #10 / 14 | ↓13.04 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)sexual_content | #11 / 14 | ↓12.88 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)specialized_advice | #13 / 14 | ↓12.67 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)suicide_self_harm | #13 / 14 | ↓15.15 | Source ↗self run | |
| Adversarial Poetry Refusal (AHB self-run)violent_crimes | #11 / 14 | ↓18.94 | Source ↗self run | |
| SpeciEvalbelief_animal_sentience | #92 / 131 | ↑6.63 | Source ↗official | |
| SpeciEvalland_animal_4ns | #6 / 131 | ↓3.72 | Source ↗official | |
| SpeciEvalsea_animal_4ns | #5 / 131 | ↓4.08 | Source ↗official | |
| SpeciEvalspeciesism | #90 / 131 | ↓2.23 | Source ↗official |
Freedom evals
Rank and direction are specific to the Freedom portfolio. Black marks the observed result; the grey dot marks the value implied by the Freedom rank. Values stay on each source’s native scale.