Our new X account is live! Follow @wizwand_team for updates

Jailbreak Defense on GCG

1Harmful Score

Retokenization

Updated 4d ago

Evaluation Results

Method	Links
Retokenization 2024.02		1	2
Self-Reminder 2024.02		1	0
ICD 2024.02		1	0
SafeDecoding 2024.02		1	0
Goal Prioritization 2025.05		1	0
IA 2025.05		1	0
SAGE 2025.05		1	0
Self-Reminder 2025.05		1	2
Self-Examination 2025.05		1	0
Goal Prioritization 2025.05		1	0
SAGE 2025.05		1	0
ICD 2025.05		1	0
SAGE 2025.05		1	0
PPL 2024.02		1.02	0
PPL 2024.02		1.06	0
Paraphrase 2024.02		1.06	4
Self-Examination 2025.05		1.06	6
Self-Examination 2025.05		1.08	0
IA 2025.05		1.08	2
Self-Reminder 2025.05		1.1	4
SafeDecoding 2024.02		1.12	4
ICD 2025.05		1.18	6
IA 2025.05		1.18	7
ICD 2025.05		1.24	6
Self-Reminder 2025.05		1.24	12
Goal Prioritization 2025.05		1.36	28
Self-Examination 2024.02		1.4	12
Self-Examination 2024.02		1.56	12
Retokenization 2024.02		1.58	42
Paraphrase 2024.02		1.8	20
No Defense 2025.05		1.88	22
No Defense 2024.02		2.48	32
Self-Reminder 2024.02		2.76	42
No Defense 2025.05		3	60
ICD 2024.02		3.86	70
No Defense 2024.02		4.7	100
No Defense 2025.05		4.74	96