Skip to content

Benchmark

HELM AIR-Bench: AIRBench 2024 - Criminal Activities

Safety benchmarks examine risks in model behaviour, including how models respond to unsafe requests. Scores are percentages. Higher is better.

Models measured
70
Scores recorded
75
Sources
1
IQ weight
1.73

Score distribution

78.8%
Mean
89.2%
Median
24.9%
Std dev
31.7%
10th percentile
99.2%
90th percentile
3.3%
Min
3.3%100%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.73 for chat models, measured over 70 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51GLM 4.5 Air· Stanford HELM specialized suites data docs79.2%
52GPT 4osetting not named by the source· Stanford HELM specialized suites data docs76.7%
53Qwen2.5 72B Instruct Turbo· Stanford HELM specialized suites data docs75.8%
54Yi 34B Chat· Stanford HELM specialized suites data docs74.6%
55GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs73.3%
56Llama 4 Scout· Stanford HELM specialized suites data docs72.5%
57DeepSeek LLM 67B Chat· Stanford HELM specialized suites data docs70%
58Qwen1.5-72B-Chat· Stanford HELM specialized suites data docs65%
59DeepSeek R1· Stanford HELM specialized suites data docs62.9%
60Qwen2.5 7B Instruct Turbo· Stanford HELM specialized suites data docs62.5%
61Grok 3 Mini· Stanford HELM specialized suites data docs59.2%
62DeepSeek V3· Stanford HELM specialized suites data docs58.3%
63Meta Llama 3.1 70B Instruct Turbo· Stanford HELM specialized suites data docs58.3%
64GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs54.2%
65Grok 3· Stanford HELM specialized suites data docs51.7%
66Grok 4· Stanford HELM specialized suites data docs50.8%
67Mixtral 8x22B Instruct v0.1· Stanford HELM specialized suites data docs38.8%
68Mistral Large2407· Stanford HELM specialized suites data docs31.7%
69Command R+· Stanford HELM specialized suites data docs28.3%
70O1 Mini· Stanford HELM specialized suites data docs28.3%
71Mixtral 8x7B Instruct v0.1· Stanford HELM specialized suites data docs25.8%
72Mistral Small· Stanford HELM specialized suites data docs20%
73Command R Plus· Stanford HELM specialized suites data docs18.3%
74Mistral (7B) Instruct v0.3· Stanford HELM specialized suites data docs5.8%
75DBRX-instruct· Stanford HELM specialized suites data docs3.3%

Showing 51–75 of 75 results.

Data credits: Stanford HELM (Stanford CRFM); scenario scores from Capabilities, Safety, Long Context, MedHELM, AIR-Bench, MMLU and Finance