Model profile
Qwen1.5 4B Chat
Evidence summary
Safety. Qwen1.5 4B Chat does not meet the evidence gate for a Safety rank. Published Safety evidence spans 2 eval lineages and 4 of 7 components. Its strongest relative result is SALAD-Bench (base_misinformation_harms, #7 of 33); its weakest is SALAD-Bench (mcq_representation_toxicity, #31 of 33).
Freedom. Qwen1.5 4B Chat has an estimated Freedom rank of #426; its 90% source-sensitivity interval is #245–#549. Published Freedom evidence spans 2 eval lineages and 1 of 1 components. Its strongest relative result is SORRY-Bench (avg, #28 of 51); its weakest is SALAD-Bench (base_misinformation_harms, #27 of 33).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗Qwen/Qwen1.5-4B-Chat
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗direct research preview identity date
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 |
|---|---|---|---|---|
| SALAD-Benchattack_enhanced_human_autonomy_integrity | #8 / 33 | ↑26.29 | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #10 / 33 | ↑23.13 | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #10 / 33 | ↑22.84 | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #7 / 33 | ↑28.29 | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #11 / 33 | ↑22.09 | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #13 / 33 | ↑19.05 | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #11 / 33 | ↑96.62 | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #13 / 33 | ↑96.14 | Source ↗official | |
| SALAD-Benchbase_malicious_use | #9 / 33 | ↑96.93 | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #7 / 33 | ↑96.45 | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #10 / 33 | ↑93.05 | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #7 / 33 | ↑94.01 | Source ↗official | |
| SALAD-Benchmcq_human_autonomy_integrity | #30 / 33 | ↑0 | Source ↗official | |
| SALAD-Benchmcq_information_safety_harms | #30 / 33 | ↑0 | Source ↗official | |
| SALAD-Benchmcq_malicious_use | #31 / 33 | ↑0 | Source ↗official | |
| SALAD-Benchmcq_misinformation_harms | #30 / 33 | ↑0 | Source ↗official | |
| SALAD-Benchmcq_representation_toxicity | #31 / 33 | ↑0 | Source ↗official | |
| SALAD-Benchmcq_socioeconomic_harms | #30 / 33 | ↑0 | Source ↗official | |
| SORRY-Benchavg | #24 / 51 | ↓0.27 | 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 |
|---|---|---|---|---|
| SALAD-Benchattack_enhanced_human_autonomy_integrity | #26 / 33 | ↓26.29 | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #24 / 33 | ↓23.13 | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #24 / 33 | ↓22.84 | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #27 / 33 | ↓28.29 | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #23 / 33 | ↓22.09 | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #21 / 33 | ↓19.05 | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #23 / 33 | ↓96.62 | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #21 / 33 | ↓96.14 | Source ↗official | |
| SALAD-Benchbase_malicious_use | #25 / 33 | ↓96.93 | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #27 / 33 | ↓96.45 | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #24 / 33 | ↓93.05 | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #27 / 33 | ↓94.01 | Source ↗official | |
| SORRY-Benchavg | #28 / 51 | ↑0.27 | Source ↗official |