Model profile
Llama 2 Chinese 13B Chat
2023-07-24release date
2eval lineages
Evidence summary
Published evidence spans 2 evals and 3 of 7 behavior components. Its strongest relative result is CMoralEval (familial_morality, #9 of 26); its weakest is SafetyBench (OFF, #21 of 21).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗FlagAlpha/Llama2-Chinese-13b-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 |
|---|---|---|---|---|---|
| CMoralEvalfamilial_morality | #9 / 26 | 0.46 | ↑ higher | Source ↗official | |
| CMoralEvalinternet_ethics | #9 / 26 | 0.45 | ↑ higher | Source ↗official | |
| CMoralEvalpersonal_morality | #9 / 26 | 0.45 | ↑ higher | Source ↗official | |
| CMoralEvalprofessional_ethics | #10 / 26 | 0.45 | ↑ higher | Source ↗official | |
| CMoralEvalsocial_morality | #10 / 26 | 0.45 | ↑ higher | Source ↗official | |
| SafetyBenchEM | #18 / 21 | 52.3 | ↑ higher | Source ↗official | |
| SafetyBenchIA | #16 / 21 | 66.9 | ↑ higher | Source ↗official | |
| SafetyBenchMH | #19 / 21 | 69.4 | ↑ higher | Source ↗official | |
| SafetyBenchOFF | #21 / 21 | 48.1 | ↑ higher | Source ↗official | |
| SafetyBenchPH | #20 / 21 | 49.7 | ↑ higher | Source ↗official | |
| SafetyBenchPP | #19 / 21 | 64.7 | ↑ higher | Source ↗official | |
| SafetyBenchUB | #17 / 21 | 54.4 | ↑ higher | Source ↗official |