Model profile
Pythia 12B Deduped
Evidence summary
Safety. Pythia 12B Deduped does not meet the evidence gate for a Safety rank. Published Safety evidence spans 1 eval lineages and 1 of 7 components. Its strongest relative result is MedSafetyBench (medical_safety_score, #17 of 30); its weakest is MedSafetyBench (medical_safety_score, #17 of 30).
Freedom. Pythia 12B Deduped has an estimated Freedom rank of #266; its 90% source-sensitivity interval is #79–#528. Published Freedom evidence spans 1 eval lineages and 1 of 1 components. Its strongest relative result is MedSafetyBench (medical_safety_score, #14 of 30); its weakest is MedSafetyBench (medical_safety_score, #14 of 30).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗EleutherAI/pythia-12b-deduped
- Release source ↗repository created at proxy
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 |
|---|---|---|---|---|
| MedSafetyBenchmedical_safety_score | #17 / 30 | ↑77.36 | 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 |
|---|---|---|---|---|
| MedSafetyBenchmedical_safety_score | #14 / 30 | ↓77.36 | Source ↗official |