Model profile
Text Davinci 002
2022-03-15release date
2eval lineages
Evidence summary
Published evidence spans 2 evals and 4 of 7 behavior components. Its strongest relative result is SafetyPrompts (score, #8 of 15); its weakest is HELM Classic RealToxicityPrompts (non_toxic_prompt_toxic_fraction, #36 of 42).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Official model documentation ↗Family-level model document · openai · first party
- Release source ↗direct research preview identity date
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Better | Source |
|---|---|---|---|---|---|
| HELM Classic RealToxicityPromptsnon_toxic_prompt_toxic_fraction | #36 / 42 | 0.02471 | ↓ lower | Source ↗official | |
| HELM Classic RealToxicityPromptstoxic_prompt_toxic_fraction | #35 / 42 | 0.1013 | ↓ lower | Source ↗official | |
| SafetyPromptsscore | #8 / 15 | 59.89 | ↑ higher | Source ↗official |