Model profile
Vicuna 13B V1.5 16K
2023-03-30release date
1eval lineages
Evidence summary
Published evidence spans 1 evals and 4 of 7 behavior components. Its strongest relative result is SALAD-Bench (mcq_socioeconomic_harms, #12 of 33); its weakest is SALAD-Bench (base_information_safety_harms, #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-13b-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 | #31 / 33 | 4.53 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #33 / 33 | 2.93 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #31 / 33 | 4.24 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #33 / 33 | 3.29 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #32 / 33 | 3.19 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #32 / 33 | 5.19 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #32 / 33 | 47.35 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #33 / 33 | 57.28 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_malicious_use | #32 / 33 | 34.59 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #32 / 33 | 62.78 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #32 / 33 | 51.39 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #32 / 33 | 57.58 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_human_autonomy_integrity | #16 / 33 | 40.28 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_information_safety_harms | #13 / 33 | 37.5 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_malicious_use | #16 / 33 | 38.21 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_misinformation_harms | #13 / 33 | 40.24 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_representation_toxicity | #13 / 33 | 40.21 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_socioeconomic_harms | #12 / 33 | 40.56 | ↑ higher | Source ↗official |