Model profile
Internlm2 Chat 7B
InternLMdeveloper
2024-01-17release date
#52 / 267overall rank
4eval lineages
Evidence summary
Internlm2 Chat 7B has an estimated overall rank of #52; its 90% source-sensitivity interval is #14–#221. Its behavior-only rank is #54; company governance moves the combined estimate to #52. Published evidence spans 4 evals and 4 of 7 behavior components. Its strongest relative result is SALAD-Bench (base_misinformation_harms, #2 of 33); its weakest is ChineseSafe (score, #17 of 22).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗internlm/internlm2-chat-7b
- 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 | #17 / 22 | 49.49 | ↑ higher | Source ↗official | |
| CMoralEvalfamilial_morality | #5 / 26 | 0.51 | ↑ higher | Source ↗official | |
| CMoralEvalinternet_ethics | #4 / 26 | 0.51 | ↑ higher | Source ↗official | |
| CMoralEvalpersonal_morality | #4 / 26 | 0.5 | ↑ higher | Source ↗official | |
| CMoralEvalprofessional_ethics | #4 / 26 | 0.52 | ↑ higher | Source ↗official | |
| CMoralEvalsocial_morality | #4 / 26 | 0.51 | ↑ higher | Source ↗official | |
| JailBenchjailbreak_success_rate | #5 / 14 | 51.22 | ↓ lower | Source ↗official | |
| SALAD-Benchattack_enhanced_human_autonomy_integrity | #11 / 33 | 21.12 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #14 / 33 | 17.59 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #8 / 33 | 24.31 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #16 / 33 | 16.45 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #13 / 33 | 19.27 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #12 / 33 | 20.35 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #7 / 33 | 98.02 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #18 / 33 | 94.45 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_malicious_use | #3 / 33 | 98.63 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #2 / 33 | 98.67 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #3 / 33 | 97.19 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #5 / 33 | 94.71 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_human_autonomy_integrity | #5 / 33 | 63.89 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_information_safety_harms | #5 / 33 | 57.5 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_malicious_use | #5 / 33 | 62.24 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_misinformation_harms | #5 / 33 | 62.62 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_representation_toxicity | #5 / 33 | 61.25 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_socioeconomic_harms | #5 / 33 | 56.11 | ↑ higher | Source ↗official |