Share your thoughts, 1 month free Claude Pro on usSee more

Jailbreaking on EHRAgent TREQS

71.78SR

Non-attack

Updated 11d ago

Evaluation Results

Method	Links
Non-attack 2026.04		71.78	-	-	-	92.23
JailAgent 2026.04		70.86	63.91	62.47	61.25	93.05
BadChain 2026.04		66.16	54.5	51.33	50.92	88.85
AgentPoison 2026.04		61.96	53.48	51.43	49.08	86.4
JailAgent 2026.04		57.87	71.88	69.02	66.97	77.2
Non-attack 2026.04		57.36	-	-	-	77.81
PAIR 2026.04		55.93	46.52	44.27	42.13	80.67
AgentPoison 2026.04		54.81	72.8	69.22	65.85	73.93
BadChain 2026.04		53.17	65.03	61.35	60.12	71.68
JailAgent 2026.04		50.1	64.83	62.58	59.2	60.94
Non-attack 2026.04		50	-	-	-	71.57
Non-attack 2026.04		49.9	-	-	-	60.63
PAIR 2026.04		47.96	53.37	51.02	49.59	65.34
JailAgent 2026.04		47.34	68.1	63.7	61.55	71.78
AgentPoison 2026.04		46.83	60.94	56.75	53.37	65.85
BadChain 2026.04		45.4	64.52	62.47	60.94	63.7
BadChain 2026.04		44.58	61.76	59	53.8	54.81
Non-attack 2026.04		44.58	-	-	-	82.31
JailAgent 2026.04		43.46	52.04	46.42	45.91	82.11
AgentPoison 2026.04		43.15	60.53	55.93	56.13	53.78
PAIR 2026.04		42.94	55.11	56.03	49.9	61.55
AgentPoison 2026.04		41.41	51.02	46.52	45.4	79.75
PAIR 2026.04		40.59	58.18	54.6	51.23	51.33
Non-attack 2026.04		40.18	-	-	-	99.49
BadChain 2026.04		39.78	47.96	44.27	42.02	76.58
JailAgent 2026.04		39.78	54.81	51.74	51.53	99.18
AgentPoison 2026.04		38.34	49.9	46.63	44.58	98.06
PAIR 2026.04		36.4	43.05	39.98	39.67	71.88
BadChain 2026.04		36.4	47.55	44.17	43.25	96.63
PAIR 2026.04		33.23	44.89	43.15	42.43	92.02