Model profile
Phi 3 Mini
1eval lineages
Evidence summary
Published evidence spans 1 evals and 2 of 7 behavior components. Its strongest relative result is Microsoft Phi Safety Panels (third_party_harm, #2 of 10); its weakest is Microsoft Phi Safety Panels (jailbreak, #6 of 10).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗phi-3-mini
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 |
|---|---|---|---|---|---|
| Microsoft Phi Safety Panelsharmful_continuation | #4 / 10 | 0.007 | ↓ lower | Source ↗official | |
| Microsoft Phi Safety Panelsharmful_summarization | #2 / 10 | 0.1 | ↓ lower | Source ↗official | |
| Microsoft Phi Safety Panelsjailbreak | #6 / 10 | 0.123 | ↓ lower | Source ↗official | |
| Microsoft Phi Safety Panelsthird_party_harm | #2 / 10 | 0.24 | ↓ lower | Source ↗official |