Model profile
Qwen1.5 14B Chat
Alibabadeveloper
2024-02-04release date
#141 / 267overall rank
6eval lineages
Evidence summary
Qwen1.5 14B Chat has an estimated overall rank of #141; its 90% source-sensitivity interval is #57–#198. Its behavior-only rank is #134; company governance moves the combined estimate to #141. Published evidence spans 6 evals and 5 of 7 behavior components. Its strongest relative result is CRiskEval (malicious_coordination, #2 of 17); its weakest is SALAD-Bench (attack_enhanced_misinformation_harms, #31 of 33).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗Qwen/Qwen1.5-14B-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 | Better | Source |
|---|---|---|---|---|---|
| ChineseSafescore | #7 / 22 | 61.29 | ↑ higher | Source ↗official | |
| ChiSafetyBenchharmful_response_rate | #2 / 14 | 0.22 | ↓ lower | Source ↗official | |
| ChiSafetyBenchmcq_score | #3 / 12 | 88.39 | ↑ higher | Source ↗official | |
| CRiskEvaldeception_willingness | #11 / 17 | 23.53 | ↓ lower | Source ↗official | |
| CRiskEvaldesire_for_resource | #5 / 17 | 28.02 | ↓ lower | Source ↗official | |
| CRiskEvalharmful_goal | #5 / 17 | 34.36 | ↓ lower | Source ↗official | |
| CRiskEvalimprovement_intent | #5 / 17 | 46.04 | ↓ lower | Source ↗official | |
| CRiskEvalmalicious_coordination | #2 / 17 | 6.52 | ↓ lower | Source ↗official | |
| CRiskEvalself_preservation | #5 / 17 | 29.41 | ↓ lower | Source ↗official | |
| CRiskEvalsituational_awareness | #12 / 17 | 66.62 | ↓ lower | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #97 / 260 | 19.12 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #165 / 260 | 86.5 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #166 / 260 | 57.78 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #237 / 258 | 80.95 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_human_autonomy_integrity | #26 / 33 | 7.11 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #30 / 33 | 3.91 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #26 / 33 | 6.53 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #31 / 33 | 4.28 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #26 / 33 | 6.19 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #22 / 33 | 9.09 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #14 / 33 | 96.1 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #8 / 33 | 97.77 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_malicious_use | #10 / 33 | 96.87 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #10 / 33 | 95.72 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #11 / 33 | 92.71 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #8 / 33 | 93.65 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_human_autonomy_integrity | #6 / 33 | 62.22 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_information_safety_harms | #9 / 33 | 49.72 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_malicious_use | #7 / 33 | 55.77 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_misinformation_harms | #8 / 33 | 54.52 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_representation_toxicity | #7 / 33 | 57.71 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_socioeconomic_harms | #7 / 33 | 52.78 | ↑ higher | Source ↗official | |
| SORRY-Benchavg | #30 / 51 | 0.34 | ↓ lower | Source ↗official |