Model profile
DeepSeek R1 Distill Llama 70B
Evidence summary
Safety. DeepSeek R1 Distill Llama 70B has an estimated Safety rank of #325; its 90% source-sensitivity interval is #275–#329. Its behavior-only rank is #315; company governance moves the combined estimate to #325. Published Safety evidence spans 4 eval lineages and 3 of 7 components. Its strongest relative result is AA-Omniscience (hallucination_rate, #173 of 345); its weakest is SOSBench (chemistry_pvr, #23 of 23).
Freedom. DeepSeek R1 Distill Llama 70B has an estimated Freedom rank of #32; its 90% source-sensitivity interval is #7–#496. Published Freedom evidence spans 2 eval lineages and 1 of 1 components. Its strongest relative result is SOSBench (chemistry_pvr, #1 of 23); its weakest is PandaBench JBB direct-request panel (safety_rate, #14 of 46).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗deepseek-r1-distill-llama-70b
- Hugging Face ↗deepseek-ai/DeepSeek-R1-Distill-Llama-70B
- OpenRouter ↗deepseek/deepseek-r1-distill-llama-70b
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗exact first party family release
Finetunes
Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.
huihui-ai/DeepSeek-R1-Distill-Llama-70B-abliterated ↗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 | #173 / 345 | ↓0.8109 | Source ↗official | |
| AbstentionBenchanswer_unknown_f1 | #17 / 20 | ↑0.7501 | Source ↗official | |
| AbstentionBenchfalse_premise_f1 | #16 / 20 | ↑0.6222 | Source ↗official | |
| AbstentionBenchstale_f1 | #19 / 20 | ↑0.5268 | Source ↗official | |
| AbstentionBenchsubjective_f1 | #19 / 20 | ↑0.525 | Source ↗official | |
| AbstentionBenchunderspecified_context_f1 | #13 / 20 | ↑0.5202 | Source ↗official | |
| AbstentionBenchunderspecified_intent_f1 | #17 / 20 | ↑0.6276 | Source ↗official | |
| PandaBench JBB direct-request panelsafety_rate | #28 / 46 | ↑0.97 | Source ↗official | |
| SOSBenchbiology_pvr | #20 / 23 | ↓0.838 | Source ↗official | |
| SOSBenchchemistry_pvr | #23 / 23 | ↓0.904 | Source ↗official | |
| SOSBenchmedicine_pvr | #20 / 23 | ↓0.854 | Source ↗official | |
| SOSBenchpharmacology_pvr | #23 / 23 | ↓0.972 | Source ↗official | |
| SOSBenchphysics_pvr | #23 / 23 | ↓0.886 | Source ↗official | |
| SOSBenchpsychology_pvr | #20 / 23 | ↓0.816 | 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.
| Eval / sub-eval | Rank | Value | Distribution | Source |
|---|---|---|---|---|
| PandaBench JBB direct-request panelsafety_rate | #14 / 46 | ↓0.97 | Source ↗official | |
| SOSBenchbiology_pvr | #4 / 23 | ↑0.838 | Source ↗official | |
| SOSBenchchemistry_pvr | #1 / 23 | ↑0.904 | Source ↗official | |
| SOSBenchmedicine_pvr | #4 / 23 | ↑0.854 | Source ↗official | |
| SOSBenchpharmacology_pvr | #1 / 23 | ↑0.972 | Source ↗official | |
| SOSBenchphysics_pvr | #1 / 23 | ↑0.886 | Source ↗official | |
| SOSBenchpsychology_pvr | #4 / 23 | ↑0.816 | Source ↗official |