Model profile
Llama 3.2 1B Instruct
Evidence summary
Safety. Llama 3.2 1B Instruct has an estimated Safety rank of #299; its 90% source-sensitivity interval is #230–#322. Its behavior-only rank is #290; company governance moves the combined estimate to #299. Published Safety evidence spans 9 eval lineages and 6 of 7 components. Its strongest relative result is Enkrypt AI Safety Leaderboard (toxicity_attack_non_success_rate, #47 of 246); its weakest is ThaiSafetyBench (safety_score, #18 of 18).
Freedom. Llama 3.2 1B Instruct has an estimated Freedom rank of #174; its 90% source-sensitivity interval is #104–#494. Published Freedom evidence spans 7 eval lineages and 1 of 1 components. Its strongest relative result is ThaiSafetyBench (safety_score, #1 of 18); its weakest is Enkrypt AI Safety Leaderboard (toxicity_attack_non_success_rate, #200 of 246).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Artificial Analysis ↗llama-3-2-instruct-1b
- Hugging Face ↗meta-llama/Llama-3.2-1B-Instruct
- NVIDIA ↗qc69jvmznzxy/llama-3.2-1b-instruct
- OpenRouter ↗meta-llama/llama-3.2-1b-instruct
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗direct research preview identity date
Finetunes
Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.
- cazzz307/Abliterated-Llama-3.2-1B-Instruct ↗Direct non-quantized finetune documented by Hugging Face metadata
- yethdev/llama3.2-1b-manumit-v1 ↗Direct non-quantized finetune documented by Hugging Face metadata
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 |
|---|---|---|---|---|
| AA-Omnisciencehallucination_rate | #113 / 345 | ↓0.6634 | Source ↗official | |
| AgentDrive Safety Compliancescr | #47 / 48 | ↑40 | Source ↗official | |
| BlueBench AttaQ-100attaq_harmlessness_reward_pct | #14 / 18 | ↑83 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #38 / 104 | ↓21.39 | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #239 / 248 | ↑4.13 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #118 / 248 | ↑88.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #50 / 248 | ↑91.11 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #47 / 246 | ↑98.82 | Source ↗official | |
| Large-scale Moral Machine experiment on LLMshuman_choice_distance | #25 / 39 | ↓1.182 | Source ↗official | |
| Open LLM Safety Indexjailbreakbench_safety_rate | #14 / 21 | ↑0.2667 | Source ↗official | |
| Open LLM Safety Indexstrongreject_safety_rate | #14 / 21 | ↑0.1333 | Source ↗official | |
| PandaBench JBB direct-request panelsafety_rate | #14 / 46 | ↑0.99 | Source ↗official | |
| ThaiSafetyBenchsafety_score | #18 / 18 | ↑62.34 | 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 |
|---|---|---|---|---|
| BlueBench AttaQ-100attaq_harmlessness_reward_pct | #5 / 18 | ↓83 | Source ↗official | |
| Cisco AI Defense Rolling Single-Turn Leaderboardsingle_turn_attack_success_rate | #67 / 104 | ↑21.39 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #127 / 248 | ↓88.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #196 / 248 | ↓91.11 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #200 / 246 | ↓98.82 | Source ↗official | |
| Open LLM Safety Indexjailbreakbench_safety_rate | #7 / 21 | ↓0.2667 | Source ↗official | |
| Open LLM Safety Indexstrongreject_safety_rate | #7 / 21 | ↓0.1333 | Source ↗official | |
| PandaBench JBB direct-request panelsafety_rate | #24 / 46 | ↓0.99 | Source ↗official | |
| ThaiSafetyBenchsafety_score | #1 / 18 | ↓62.34 | Source ↗official | |
| UGI Leaderboard — base-model willingnesswillingness_adherence_score | #126 / 156 | ↑1 | Source ↗official | |
| UGI Leaderboard — base-model willingnesswillingness_direct_score | #113 / 156 | ↑2 | 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.