Model profile
Text Davinci 003
Evidence summary
Safety. Text Davinci 003 has an estimated Safety rank of #169; its 90% source-sensitivity interval is #70–#252. Its behavior-only rank is #182; company governance moves the combined estimate to #169. Published Safety evidence spans 3 eval lineages and 4 of 7 components. Its strongest relative result is SafetyBench (OFF, #4 of 21); its weakest is SafetyBench (UB, #14 of 21).
Freedom. Text Davinci 003 has an estimated Freedom rank of #384; its 90% source-sensitivity interval is #166–#579. Published Freedom evidence spans 2 eval lineages and 1 of 1 components. Its strongest relative result is SafetyPrompts (score, #9 of 15); its weakest is HELM Classic RealToxicityPrompts (non_toxic_prompt_toxic_fraction, #31 of 40).
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
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 |
|---|---|---|---|---|
| HELM Classic RealToxicityPromptsnon_toxic_prompt_toxic_fraction | #10 / 40 | ↓0.01004 | Source ↗official | |
| HELM Classic RealToxicityPromptstoxic_prompt_toxic_fraction | #12 / 40 | ↓0.06726 | Source ↗official | |
| SafetyBenchEM | #7 / 21 | ↑72.95 | Source ↗official | |
| SafetyBenchIA | #6 / 21 | ↑81.8 | Source ↗official | |
| SafetyBenchMH | #9 / 21 | ↑82.35 | Source ↗official | |
| SafetyBenchOFF | #4 / 21 | ↑73.2 | Source ↗official | |
| SafetyBenchPH | #6 / 21 | ↑74.8 | Source ↗official | |
| SafetyBenchPP | #6 / 21 | ↑80.2 | Source ↗official | |
| SafetyBenchUB | #14 / 21 | ↑60.45 | Source ↗official | |
| SafetyPromptsscore | #7 / 15 | ↑84.76 | 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 |
|---|---|---|---|---|
| HELM Classic RealToxicityPromptsnon_toxic_prompt_toxic_fraction | #31 / 40 | ↑0.01004 | Source ↗official | |
| HELM Classic RealToxicityPromptstoxic_prompt_toxic_fraction | #29 / 40 | ↑0.06726 | Source ↗official | |
| SafetyPromptsscore | #9 / 15 | ↓84.76 | Source ↗official |