List your company

Jailbreaking on EHRAgent ALL

70.112Weighted Average ASRJailAgent
48.0858453.8041759.522565.24083Apr 7, 2026

Evaluation results

MethodLinks
JailAgentAgent Backbone=GPT-4o2026.0470.112
JailAgentAgent Backbone=Claude-3.5-haiku2026.0469.336
JailAgentAgent Backbone=Gemini-3.0-pro2026.0468.466
JailAgentAgent Backbone=GPT-52026.0466.753
AgentPoisonAgent Backbone=GPT-4o2026.0466.708
AgentPoisonAgent Backbone=Claude-3.5-haiku2026.0465.763
BadChainAgent Backbone=Claude-3.5-haiku2026.0463.639
BadChainAgent Backbone=GPT-4o2026.0463.555
BadChainAgent Backbone=GPT-52026.0462.619
AgentPoisonAgent Backbone=GPT-52026.0462.142
JailAgentAgent Backbone=GPT-3.5-turbo2026.0462.057
BadChainAgent Backbone=Gemini-3.0-pro2026.0461.927
JailAgentAgent Backbone=Llama-3.1-70B2026.0461.291
AgentPoisonAgent Backbone=Gemini-3.0-pro2026.0461.15
PAIRAgent Backbone=Claude-3.5-haiku2026.0458.559
AgentPoisonAgent Backbone=Llama-3.1-70B2026.0458.54
BadChainAgent Backbone=Llama-3.1-70B2026.0458.213
BadChainAgent Backbone=GPT-3.5-turbo2026.0456.729
AgentPoisonAgent Backbone=GPT-3.5-turbo2026.0456.136
PAIRAgent Backbone=GPT-52026.0455.686
PAIRAgent Backbone=GPT-4o2026.0454.379
PAIRAgent Backbone=Gemini-3.0-pro2026.0453.854
PAIRAgent Backbone=Llama-3.1-70B2026.0451.347
PAIRAgent Backbone=GPT-3.5-turbo2026.0448.933