Model profile
Internlm2 Chat 20B
InternLMdeveloper
2024-01-17release date
#59 / 267overall rank
4eval lineages
Evidence summary
Internlm2 Chat 20B has an estimated overall rank of #59; its 90% source-sensitivity interval is #21–#222. Published evidence spans 4 evals and 5 of 7 behavior components. Its strongest relative result is SALAD-Bench (base_representation_toxicity, #2 of 33); its weakest is ChineseSafe (score, #15 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-20b
- 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 | #15 / 22 | 53.67 | ↑ higher | Source ↗official | |
| CMoralEvalfamilial_morality | #3 / 26 | 0.56 | ↑ higher | Source ↗official | |
| CMoralEvalinternet_ethics | #3 / 26 | 0.54 | ↑ higher | Source ↗official | |
| CMoralEvalpersonal_morality | #3 / 26 | 0.52 | ↑ higher | Source ↗official | |
| CMoralEvalprofessional_ethics | #3 / 26 | 0.54 | ↑ higher | Source ↗official | |
| CMoralEvalsocial_morality | #3 / 26 | 0.54 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #88 / 260 | 20.41 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #41 / 260 | 93.33 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #128 / 260 | 71.67 | ↑ higher | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #127 / 258 | 96.41 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_human_autonomy_integrity | #15 / 33 | 15.52 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_information_safety_harms | #21 / 33 | 7.17 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_malicious_use | #15 / 33 | 12.56 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_misinformation_harms | #19 / 33 | 9.38 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_representation_toxicity | #20 / 33 | 10.12 | ↑ higher | Source ↗official | |
| SALAD-Benchattack_enhanced_socioeconomic_harms | #17 / 33 | 12.55 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_human_autonomy_integrity | #3 / 33 | 98.54 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_information_safety_harms | #11 / 33 | 96.68 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_malicious_use | #2 / 33 | 98.88 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_misinformation_harms | #2 / 33 | 98.67 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_representation_toxicity | #2 / 33 | 97.53 | ↑ higher | Source ↗official | |
| SALAD-Benchbase_socioeconomic_harms | #3 / 33 | 95.77 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_human_autonomy_integrity | #8 / 33 | 61.11 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_information_safety_harms | #7 / 33 | 51.11 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_malicious_use | #6 / 33 | 57.24 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_misinformation_harms | #6 / 33 | 59.05 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_representation_toxicity | #6 / 33 | 57.92 | ↑ higher | Source ↗official | |
| SALAD-Benchmcq_socioeconomic_harms | #10 / 33 | 46.11 | ↑ higher | Source ↗official |