Jailbreak Defense on ReNeLLM

0ASR

Updated 1mo ago

Evaluation Results

Method
IA 2025.05	0	1
SAGE 2025.05	0	1
Goal Prioritization 2025.05	0	1
SAGE 2025.05	0	1
ICD 2025.05	0	1.02
Goal Prioritization 2025.05	0	1
SAGE 2025.05	0	1
SAGE 2025.05	0	1
IA 2025.05	0	1
IA 2025.05	0.04	1.08
Self-Reminder 2025.05	0.06	1.08
Self-Examination 2025.05	0.12	1.34
IA 2025.05	0.14	1.06
Self-Examination 2025.05	0.16	1.62
No Defense 2025.05	0.2	1.72
Self-Examination 2025.05	0.22	1.76
Goal Prioritization 2025.05	0.24	1.64
ICD 2025.05	0.24	1
ICD 2025.05	0.36	2.32
Self-Reminder 2025.05	0.52	2.02
Self-Reminder 2025.05	0.6	1.7
No Defense 2025.05	0.98	4.74
No Defense 2025.05	1	4.74
Self-Examination 2025.05	2	1.08
SAGE 2025.05	2	1
SAGE 2025.05	6	1.16
Goal Prioritization 2025.05	8	1.24
ICD 2025.05	10	1.4
FreoStream 2026.06	10.1	-
w/o Future-Aware Reasoning 2026.06	10.1	-
Qwen-Stream-0.6B 2026.06	11.49	-
w/o Safety-Aligned Optimization 2026.06	11.86	-
Nemotron-4B 2026.06	12.28	-
Qwen-Gen-0.6B Prompt+Response 2026.06	16.83	-
Goal Prioritization 2025.05	18	1.2
Qwen-Stream-8B 2026.06	22.01	-
YuFeng-XGuard-0.6B Prompt Only 2026.06	23.56	-
YuFeng-XGuard-0.6B Prompt+Response 2026.06	25.76	-
Self-Examination 2025.05	28	2.02
Self-Reminder 2025.05	28	1.64
GuardReasoner-1B 2026.06	28.71	-
Self-Examination 2025.05	38	1.58
Goal Prioritization 2025.05	50	1.54
No Defense 2025.05	70	3.68
Qwen-Gen-0.6B Prompt Only 2026.06	81.19	-
ICD 2025.05	88	4.2
No Defense 2025.05	96	4.76
ICD 2025.05	96	4.56
Self-Reminder 2025.05	98	3.72
IA 2025.05	98	3.14
Self-Reminder 2025.05	98	4.38
No Defense 2025.05	100	4.86
IA 2025.05	100	4.3