Model profile
Qwen 7B Chat
Alibabadeveloper
2023-08-03release date
#190 / 267overall rank
9eval lineages
Evidence summary
Qwen 7B Chat has an estimated overall rank of #190; its 90% source-sensitivity interval is #110–#214. Its behavior-only rank is #184; company governance moves the combined estimate to #190. Published evidence spans 9 evals and 4 of 7 behavior components. Its strongest relative result is CMoralEval (familial_morality, #6 of 26); its weakest is SALAD-Bench (attack_enhanced_representation_toxicity, #27 of 33).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- OpenRouter ↗qwen/qwen-7b-chat
- Official model page ↗Family-level model document · alibaba · first party
- 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 | Better | Source |
|---|---|---|---|---|---|
| CMoralEvalfamilial_morality | #6 / 26 | 0.48 | ↑ higher | Source ↗official | |
| CMoralEvalinternet_ethics | #9 / 26 | 0.45 | ↑ higher | Source ↗official | |
| CMoralEvalpersonal_morality | #7 / 26 | 0.46 | ↑ higher | Source ↗official | |
| CMoralEvalprofessional_ethics | #7 / 26 | 0.47 | ↑ higher | Source ↗official | |
| CMoralEvalsocial_morality | #7 / 26 | 0.48 | ↑ higher | Source ↗official | |
| Fake Alignment (FINE)multiple_choice_safe_decision_rate | #8 / 14 | 54.67 | ↑ higher | Source ↗official | |
| Fake Alignment (FINE)open_ended_safe_response_rate | #5 / 14 | 97.33 | ↑ higher | Source ↗official | |
| FLAMESdata_protection | #6 / 13 | 43.42 | ↑ higher | Source ↗official | |
| FLAMESfairness | #8 / 13 | 36.14 | ↑ higher | Source ↗official | |
| FLAMESlegality | #10 / 13 | 30.43 | ↑ higher | Source ↗official | |
| FLAMESmorality | #9 / 13 | 40.3 | ↑ higher | Source ↗official | |
| FLAMESsafety | #9 / 13 | 31.93 | ↑ higher | Source ↗official | |
| HarmBenchdr | #8 / 28 | 13 | ↓ lower | Source ↗official | |
| S-Evalbase_en_overall | #14 / 22 | 55.7 | ↑ higher | Source ↗official | |
| SafeDialBenchaggression | #8 / 18 | 7.127 | ↑ higher | Source ↗official | |
| SafeDialBenchethics | #5 / 18 | 7.623 | ↑ higher | Source ↗official | |
| SafeDialBenchfairness | #12 / 18 | 7.19 | ↑ higher | Source ↗official | |
| SafeDialBenchlegality | #10 / 18 | 7.93 | ↑ higher | Source ↗official | |
| SafeDialBenchmorality | #8 / 18 | 7.277 | ↑ higher | Source ↗official | |
| SafeDialBenchprivacy | #10 / 18 | 7.563 | ↑ higher | Source ↗official | |
| SafetyBenchEM | #8 / 21 | 71.4 | ↑ higher | Source ↗official | |
| SafetyBenchIA | #7 / 21 | 80.1 | ↑ higher | Source ↗official | |
| SafetyBenchMH | #5 / 21 | 84.45 | ↑ higher | Source ↗official | |
| SafetyBenchOFF | #7 / 21 | 69.1 | ↑ higher | Source ↗official | |
| SafetyBenchPH | #9 / 21 | 70.4 | ↑ higher | Source ↗official | |
| SafetyBenchPP | #8 / 21 | 77.2 | ↑ higher | Source ↗official | |
| SafetyBenchUB | #7 / 21 | 65.9 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_human_autonomy_integrity | #25 / 33 | 7.76 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #23 / 33 | 6.84 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #21 / 33 | 8.4 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #22 / 33 | 8.72 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #27 / 33 | 5.91 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #24 / 33 | 7.79 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #18 / 33 | 94.35 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #19 / 33 | 93.84 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_malicious_use | #18 / 33 | 93.24 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #19 / 33 | 93.11 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #20 / 33 | 88.39 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #18 / 33 | 88.48 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_human_autonomy_integrity | #12 / 33 | 42.5 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_information_safety_harms | #14 / 33 | 37.22 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_malicious_use | #13 / 33 | 41.35 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_misinformation_harms | #13 / 33 | 40.24 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_representation_toxicity | #14 / 33 | 40.1 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_socioeconomic_harms | #16 / 33 | 37.78 | ↑ higher | Source ↗official | |
| SuperCLUE Safetyinstruction_attack | #24 / 31 | 55.17 | ↑ higher | Source ↗official | |
| SuperCLUE Safetyresponsible_ai | #20 / 31 | 52.73 | ↑ higher | Source ↗official | |
| SuperCLUE Safetytraditional_safety | #10 / 31 | 78.72 | ↑ higher | Source ↗official |