Model profile
Llama 3.2 3B Instruct
Evidence summary
Llama 3.2 3B Instruct has an estimated overall rank of #250; its 90% source-sensitivity interval is #189–#258. Its behavior-only rank is #243; company governance moves the combined estimate to #250. Published evidence spans 8 evals and 6 of 7 behavior components. Its strongest relative result is AA-Omniscience (hallucination_rate, #66 of 311); its weakest is UAVBench safety-critical decision recognition (ethical_safety_critical_accuracy, #27 of 27).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗llama-3-2-instruct-3b
- Hugging Face ↗meta-llama/Llama-3.2-3B-Instruct
- NVIDIA ↗qc69jvmznzxy/llama-3.2-3b-instruct
- OpenRouter ↗meta-llama/llama-3.2-3b-instruct
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- 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 |
|---|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #66 / 311 | 0.5753 | ↓ lower | Source ↗official | |
| BlueBench AttaQ-100attaq_harmlessness_reward_pct | #13 / 18 | 83.37 | ↑ higher | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #58 / 105 | 40.58 | ↓ lower | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #211 / 260 | 11.11 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #177 / 260 | 85.67 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #80 / 260 | 85 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #154 / 258 | 95.5 | ↑ higher | Source ↗official | |
| Large-scale Moral Machine experiment on LLMshuman_choice_distance | #34 / 39 | 1.519 | ↓ lower | Source ↗official | |
| Open LLM Safety Indexjailbreakbench_safety_rate | #14 / 21 | 0.2667 | ↑ higher | Source ↗official | |
| Open LLM Safety Indexstrongreject_safety_rate | #5 / 21 | 0.6667 | ↑ higher | Source ↗official | |
| PandaBench JBB direct-request panelsafety_rate | #14 / 46 | 0.99 | ↑ higher | Source ↗official | |
| UAVBench safety-critical decision recognitionethical_safety_critical_accuracy | #27 / 27 | 0.475 | ↑ higher | Source ↗official |
Values evaluations
Descriptive values and political-framing results are separate from safety/ethics ranks. Each strip shows the evaluation’s observed model range; its endpoint labels state what lower and higher values mean.