Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Toxicity Mitigation on DailyDialog
Loading...
1.16
RTR (Backdoor Non-toxic)
Attack w/o defense
-0.0256
0.2822
0.59
0.8978
Jul 8, 2025
RTR (Backdoor Non-toxic)
RTR (Backdoor Trigger)
RTR (Indiscriminate Non-toxic)
RTR (Indiscriminate Toxic)
Updated 16d ago
Evaluation Results
Method
Method
Links
RTR (Backdoor Non-toxic)
RTR (Backdoor Trigger)
RTR (Indiscriminate Non-toxic)
RTR (Indiscriminate Toxic)
Attack w/o defense
Defense setting=Attack...
2025.07
1.16
62.18
13.76
54.78
Multi-level filter
Defense setting=Multi-...
2025.07
0.16
15.56
1.7
18.08
Optimus (CH)
Defense setting=Optimu...
2025.07
0.02
0.04
0.02
0.64
Feedback
Search any
task
Search any
task