Model profile
Baichuan 2 7B Chat
Baichuandeveloper
2023-09-06release date
#220 / 309overall rank
11eval lineages
Evidence summary
Baichuan 2 7B Chat has an estimated overall rank of #220; its 90% source-sensitivity interval is #155–#246. Its behavior-only rank is #224; company governance moves the combined estimate to #220. Published evidence spans 11 evals and 5 of 7 behavior components. Its strongest relative result is FLAMES (safety, #1 of 13); its weakest is CRiskEval (improvement_intent, #17 of 17).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗baichuan-inc/Baichuan2-7B-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 | Source |
|---|---|---|---|---|
| ChineseSafescore | #13 / 22 | ↑53.99 | Source ↗official | |
| ChiSafetyBenchharmful_response_rate | #8 / 14 | ↓0.65 | Source ↗official | |
| ChiSafetyBenchmcq_score | #11 / 12 | ↑50.03 | Source ↗official | |
| CMoralEvalfamilial_morality | #8 / 26 | ↑0.47 | Source ↗official | |
| CMoralEvalinternet_ethics | #6 / 26 | ↑0.48 | Source ↗official | |
| CMoralEvalpersonal_morality | #6 / 26 | ↑0.47 | Source ↗official | |
| CMoralEvalprofessional_ethics | #6 / 26 | ↑0.49 | Source ↗official | |
| CMoralEvalsocial_morality | #6 / 26 | ↑0.5 | Source ↗official | |
| CRiskEvaldeception_willingness | #16 / 17 | ↓38.4 | Source ↗official | |
| CRiskEvaldesire_for_resource | #14 / 17 | ↓39.06 | Source ↗official | |
| CRiskEvalharmful_goal | #14 / 17 | ↓52.04 | Source ↗official | |
| CRiskEvalimprovement_intent | #17 / 17 | ↓60.08 | Source ↗official | |
| CRiskEvalmalicious_coordination | #16 / 17 | ↓29.29 | Source ↗official | |
| CRiskEvalself_preservation | #13 / 17 | ↓40.2 | Source ↗official | |
| CRiskEvalsituational_awareness | #6 / 17 | ↓61.78 | Source ↗official | |
| DSPSafeBenchscore | #11 / 12 | ↑65.31 | Source ↗official | |
| Fake Alignment (FINE)multiple_choice_safe_decision_rate | #12 / 14 | ↑20 | Source ↗official | |
| Fake Alignment (FINE)open_ended_safe_response_rate | #5 / 14 | ↑97.33 | Source ↗official | |
| FLAMESdata_protection | #8 / 13 | ↑40.79 | Source ↗official | |
| FLAMESfairness | #4 / 13 | ↑42.17 | Source ↗official | |
| FLAMESlegality | #4 / 13 | ↑52.17 | Source ↗official | |
| FLAMESmorality | #11 / 13 | ↑39.3 | Source ↗official | |
| FLAMESsafety | #1 / 13 | ↑56.41 | Source ↗official | |
| HarmBenchdr | #13 / 28 | ↓18.8 | Source ↗official | |
| M3-SafetyBenchoverall_score | #16 / 19 | ↑82.91 | Source ↗official | |
| SafeDialBenchaggression | #11 / 18 | ↑7.073 | Source ↗official | |
| SafeDialBenchethics | #6 / 18 | ↑7.613 | Source ↗official | |
| SafeDialBenchfairness | #16 / 18 | ↑7.123 | Source ↗official | |
| SafeDialBenchlegality | #9 / 18 | ↑7.937 | Source ↗official | |
| SafeDialBenchmorality | #4 / 18 | ↑7.383 | Source ↗official | |
| SafeDialBenchprivacy | #11 / 18 | ↑7.523 | Source ↗official | |
| SuperCLUE Safetyinstruction_attack | #15 / 31 | ↑65.52 | Source ↗official | |
| SuperCLUE Safetyresponsible_ai | #13 / 31 | ↑61.82 | Source ↗official | |
| SuperCLUE Safetytraditional_safety | #5 / 31 | ↑80.85 | Source ↗official |