List your company

Harmful Compliance Analysis on Paired safety frame contrast items n=20 paired items per model (test)

-0.036Harm Delta (eval-dep)OLMo-3-Instruct (7B)
-0.044440.012530.06950.12647May 7, 2026

Evaluation results

MethodLinks
OLMo-3-Instruct (7B)Model family=OLMo-3, Model size (parameters)=7B, Protocol=paired-prompt protocol2026.05-0.0360.024
Llama-3.1-70BModel family=Llama, Model size (parameters)=70B, Protocol=paired-prompt protocol2026.0500.844
Phi-3.5-mini (3.8B)Model family=Phi, Model size (parameters)=3.8B, Protocol=paired-prompt protocol2026.050.0930.005
Mistral-Small-3.2 (24B)Model family=Mistral, Model size (parameters)=24B, Protocol=paired-prompt protocol2026.050.1210.003
Llama-3.1-8BModel family=Llama, Model size (parameters)=8B, Protocol=paired-prompt protocol2026.050.1750.005