Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Defending against gradient-based attacks on Llama3 GCG Attack (test)
Loading...
9.61
ASR
Ours-Fakecom-t
6.514
27.412
48.31
69.208
Nov 1, 2024
ASR
Updated 4d ago
Evaluation Results
Method
Method
Links
ASR
Ours-Fakecom-t
Defense=Fake Completio...
2024.11
9.61
Ours-Ignore
Defense=Ignore Defense
2024.11
12.01
Ours-Fakecom
Defense=Fake Completio...
2024.11
13.94
Sandwich
Defense=Sandwich defen...
2024.11
19.23
Ours-Escape
Defense=Escape Defense
2024.11
19.23
Spotlight
Defense=Spotlight defe...
2024.11
19.71
Reminder
Defense=Reminder defen...
2024.11
24.51
Instructional
Defense=Instructional...
2024.11
28.84
Isolation
Defense=Isolation defe...
2024.11
40.38
None
Defense=No defense bas...
2024.11
87.01
Feedback
Search any
task
Search any
task