Model profile
Qwen1.5 32B Chat
Alibabadeveloper
2024-04-02release date
#154 / 312overall rank
4eval lineages
Evidence summary
Qwen1.5 32B Chat has an estimated overall rank of #154; its 90% source-sensitivity interval is #84–#213. Its behavior-only rank is #149; company governance moves the combined estimate to #154. Published evidence spans 4 evals and 6 of 7 behavior components. Its strongest relative result is ChiSafetyBench (harmful_response_rate, #2 of 14); its weakest is CRiskEval (situational_awareness, #10 of 17).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗Qwen/Qwen1.5-32B-Chat
- 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 | Source |
|---|---|---|---|---|
| ChiSafetyBenchharmful_response_rate | #2 / 14 | ↓0.22 | Source ↗official | |
| ChiSafetyBenchmcq_score | #4 / 12 | ↑87.3 | Source ↗official | |
| CRiskEvaldeception_willingness | #5 / 17 | ↓19.92 | Source ↗official | |
| CRiskEvaldesire_for_resource | #6 / 17 | ↓28.56 | Source ↗official | |
| CRiskEvalharmful_goal | #7 / 17 | ↓35.65 | Source ↗official | |
| CRiskEvalimprovement_intent | #8 / 17 | ↓48.96 | Source ↗official | |
| CRiskEvalmalicious_coordination | #9 / 17 | ↓10.48 | Source ↗official | |
| CRiskEvalself_preservation | #6 / 17 | ↓29.81 | Source ↗official | |
| CRiskEvalsituational_awareness | #10 / 17 | ↓65.38 | Source ↗official | |
| OR-Benchover_refusal_rate | #13 / 25 | ↓50.8 | Source ↗official | |
| OR-Benchtoxic_acceptance_rate | #13 / 25 | ↓4.4 | Source ↗official | |
| SORRY-Benchavg | #25 / 51 | ↓0.28 | Source ↗official |