Model profile
Phi 3 Medium 128K Instruct
Evidence summary
Safety. Phi 3 Medium 128K Instruct does not meet the evidence gate for a Safety rank. Published Safety evidence spans 2 eval lineages and 3 of 7 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (toxicity_attack_non_success_rate, #106 of 246); its weakest is Enkrypt AI Safety Leaderboard (bias_attack_non_success_rate, #195 of 248).
Freedom. Phi 3 Medium 128K Instruct has an estimated Freedom rank of #298; its 90% source-sensitivity interval is #94–#572. Published Freedom evidence spans 2 eval lineages and 1 of 1 components. Its strongest relative result is SpeechMap model completion (complete_pct, #44 of 181); its weakest is Enkrypt AI Safety Leaderboard (toxicity_attack_non_success_rate, #140 of 246).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗microsoft/Phi-3-medium-128k-instruct
- OpenRouter ↗microsoft/phi-3-medium-128k-instruct
- 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 |
|---|---|---|---|---|
| AgentDrive Safety Compliancescr | #24 / 48 | ↑82.5 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #195 / 248 | ↑11.89 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #133 / 248 | ↑87.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #114 / 248 | ↑73.89 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #106 / 246 | ↑97.05 | 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 |
|---|---|---|---|---|
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #112 / 248 | ↓87.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #134 / 248 | ↓73.89 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #140 / 246 | ↓97.05 | Source ↗official | |
| SpeechMap model completioncomplete_pct | #44 / 181 | ↑75 | Source ↗official |