Model profile
Qwen3 4B
Evidence summary
Safety. Qwen3 4B has an estimated Safety rank of #102; its 90% source-sensitivity interval is #27–#240. Its behavior-only rank is #89; company governance moves the combined estimate to #102. Published Safety evidence spans 4 eval lineages and 5 of 7 components. Its strongest relative result is Vectara HHEM Factual Consistency (factual_consistency_rate, #20 of 94); its weakest is PandaBench JBB direct-request panel (safety_rate, #28 of 46).
Freedom. Qwen3 4B has an estimated Freedom rank of #118; its 90% source-sensitivity interval is #17–#357. Published Freedom evidence spans 4 eval lineages and 1 of 1 components. Its strongest relative result is UGI Leaderboard — base-model willingness (willingness_direct_score, #4 of 156); its weakest is MuPPET Contextual Privacy (multiparty_contextual_privacy_score, #5 of 7).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗Qwen/Qwen3-4B
- OpenRouter ↗qwen/qwen3-4b
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗repository created at proxy
Finetunes
Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.
huihui-ai/Huihui-Qwen3-4B-abliterated-v2 ↗Direct non-quantized finetune documented by Hugging Face metadata- huihui-ai/Qwen3-4B-abliterated ↗Direct non-quantized finetune documented by Hugging Face metadata
OBLITERATUS/Qwen3-4B-OBLITERATED ↗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 |
|---|---|---|---|---|
| MuPPET Contextual Privacymultiparty_contextual_privacy_score | #3 / 7 | ↑41.83 | Source ↗official | |
| PandaBench JBB direct-request panelsafety_rate | #28 / 46 | ↑0.97 | Source ↗official | |
| Vectara HHEM Factual Consistencyfactual_consistency_rate | #20 / 94 | ↑94.3 | Source ↗official | |
| VETO Misfired Alignmentmisfired_alignment_rate_pct | #10 / 25 | ↓7.9 | 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 |
|---|---|---|---|---|
| MuPPET Contextual Privacymultiparty_contextual_privacy_score | #5 / 7 | ↓41.83 | Source ↗official | |
| PandaBench JBB direct-request panelsafety_rate | #14 / 46 | ↓0.97 | Source ↗official | |
| UGI Leaderboard — base-model willingnesswillingness_adherence_score | #4 / 156 | ↑8.25 | Source ↗official | |
| UGI Leaderboard — base-model willingnesswillingness_direct_score | #4 / 156 | ↑6 | Source ↗official | |
| VETO Misfired Alignmentmisfired_alignment_rate_pct | #10 / 25 | ↓7.9 | 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.