Skip to content

Benchmark

HELM AIR-Bench: AIRBench 2024 - #45: Other Illegal/Unlawful/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
1.88

Score distribution

82.1%
Mean
95%
Median
25.5%
Std dev
35%
10th percentile
100%
90th percentile
0%
Min
0%100%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.88 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
51GPT 4osetting not named by the source· Stanford HELM specialized suites data docs85%
52Qwen2.5 72B Instruct Turbo· Stanford HELM specialized suites data docs85%
53gemini-1.5-flash-002· Stanford HELM specialized suites data docs83.3%
54DeepSeek LLM 67B Chat· Stanford HELM specialized suites data docs81.7%
55DeepSeek R1· Stanford HELM specialized suites data docs76.7%
56GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs76.7%
57Llama 4 Scout· Stanford HELM specialized suites data docs76.7%
58Yi 34B Chat· Stanford HELM specialized suites data docs76.7%
59Qwen1.5-72B-Chat· Stanford HELM specialized suites data docs71.7%
60DeepSeek V3· Stanford HELM specialized suites data docs68.3%
61Grok 3 Mini· Stanford HELM specialized suites data docs66.7%
62Qwen2.5 7B Instruct Turbo· Stanford HELM specialized suites data docs66.7%
63Meta Llama 3.1 70B Instruct Turbo· Stanford HELM specialized suites data docs63.3%
64GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs61.7%
65Grok 3· Stanford HELM specialized suites data docs50%
66Grok 4· Stanford HELM specialized suites data docs50%
67Mixtral 8x22B Instruct v0.1· Stanford HELM specialized suites data docs41.7%
68Mistral Large2407· Stanford HELM specialized suites data docs35%
69Command R+· Stanford HELM specialized suites data docs28.3%
70Mixtral 8x7B Instruct v0.1· Stanford HELM specialized suites data docs26.7%
71O1 Mini· Stanford HELM specialized suites data docs25%
72Command R Plus· Stanford HELM specialized suites data docs18.3%
73Mistral Small· Stanford HELM specialized suites data docs18.3%
74Mistral (7B) Instruct v0.3· Stanford HELM specialized suites data docs5%
75DBRX-instruct· Stanford HELM specialized suites data docs0%

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