Model profile
Vicuna 7B V1.5 16K
2023-06-22release date
1eval lineages
Evidence summary
Published evidence spans 1 evals and 4 of 7 behavior components. Its strongest relative result is SALAD-Bench (mcq_human_autonomy_integrity, #20 of 33); its weakest is SALAD-Bench (base_representation_toxicity, #33 of 33).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗lmsys/vicuna-7b-v1.5-16k
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗global scoring family inherited
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 |
|---|---|---|---|---|---|
| SALAD-Benchattack_enhanced_human_autonomy_integrity | #30 / 33 | 4.96 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #30 / 33 | 3.91 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #33 / 33 | 3.83 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #27 / 33 | 6.09 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #31 / 33 | 3.33 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #29 / 33 | 6.93 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #33 / 33 | 46.42 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #32 / 33 | 63.51 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_malicious_use | #33 / 33 | 33.49 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #33 / 33 | 62.04 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #33 / 33 | 47.27 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #33 / 33 | 57.11 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_human_autonomy_integrity | #20 / 33 | 27.5 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_information_safety_harms | #24 / 33 | 17.22 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_malicious_use | #21 / 33 | 23.01 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_misinformation_harms | #21 / 33 | 24.29 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_representation_toxicity | #22 / 33 | 23.65 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_socioeconomic_harms | #23 / 33 | 21.11 | ↑ higher | Source ↗official |