Model profile
Qwen3 4B Thinking
Evidence summary
Safety. Qwen3 4B Thinking does not meet the evidence gate for a Safety rank. Published Safety evidence spans 1 eval lineages and 1 of 7 components. Its strongest relative result is BrokenMath (sycophancy, #7 of 9); its weakest is BrokenMath (sycophancy, #7 of 9).
Freedom. Qwen3 4B Thinking has an estimated Freedom rank of #302; its 90% source-sensitivity interval is #132–#549. Published Freedom evidence spans 1 eval lineages and 1 of 1 components. Its strongest relative result is UGI Leaderboard — base-model willingness (willingness_direct_score, #46 of 156); its weakest is UGI Leaderboard — base-model willingness (willingness_adherence_score, #79 of 156).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗Qwen/Qwen3-4B-Thinking-2507
- Official model page ↗Family-level model document · alibaba · first party
Finetunes
Direct finetunes linked to this canonical base model. Quantized and repackaged derivatives are excluded.
heretic-org/Qwen3-4B-Thinking-2507-heretic ↗Direct non-quantized finetune documented by Hugging Face metadata
huihui-ai/Huihui-Qwen3-4B-Thinking-2507-abliterated ↗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 |
|---|---|---|---|---|
| BrokenMathsycophancy | #7 / 9 | ↓55.6 | 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 |
|---|---|---|---|---|
| UGI Leaderboard — base-model willingnesswillingness_adherence_score | #79 / 156 | ↑1.5 | Source ↗official | |
| UGI Leaderboard — base-model willingnesswillingness_direct_score | #46 / 156 | ↑4 | 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.