List your company

Fine-grained risk diagnosis on ATBench

75.2Risk Source ScoreAgentDoG 1.5-4B
-1.96818.06638.158.134May 28, 2026

Evaluation results

MethodLinks
AgentDoG 1.5-4BModel Category=Our Models, Size=4B2026.0575.227.562.955.2
AgentDoG 1.5-8BModel Category=Our Models, Size=8B2026.0572.924.652.550
AgentDoG 1.5-2BModel Category=Our Models, Size=2B2026.05682453.848.6
AgentDoG 1.5-0.8BModel Category=Our Models, Size=0.8B2026.0565.718.444.943
AgentDoG 1.0-4BModel Category=Our Models2026.0546.816.540.634.6
GPT-5.4Model Category=Closed-Source Models2026.0533.613.530.225.8
GPT-5.2Model Category=Closed-Source Models2026.0529.51226.822.8
Gemini-3.1-ProModel Category=Closed-Source Models2026.0524.812.618.518.6
AgentDoG 1.5-4B-UModel Category=Our Models, Size=4B, Type=Unified2026.0524.19.528.420.7
Gemini-3-FlashModel Category=Closed-Source Models2026.0518.48.31513.9
QwQ-32BModel Category=Open-Source Models2026.0515.89.422.916
Qwen3.5-397BModel Category=Open-Source Models2026.057.73.66.86
Qwen3.5-2BModel Category=Open-Source Models2026.057.76.611.18.5
Qwen3-235BModel Category=Open-Source Models2026.05711.626.615.1
Qwen3.5-4BModel Category=Open-Source Models2026.056.638.25.9
Llama3.1-8B-Inst.Model Category=Open-Source Models2026.056.25.815.59.2
Qwen2.5-7B-Inst.Model Category=Open-Source Models2026.055.3615.58.9
Qwen3.5-0.8BModel Category=Open-Source Models2026.051.32.94.73
Qwen3-4B-Inst.Model Category=Open-Source Models2026.0519.621.210.6