Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Lifelong Safety Adaptation on Safety Benchmarks (Final-day)
Loading...
96.2
Final-day Macro F1 Score
LiSA
81.224
85.112
89
92.888
May 14, 2026
Final-day Macro F1 Score
Updated 19d ago
Evaluation Results
Method
Method
Links
Final-day Macro F1 Score
LiSA
label-flip ratio (ρ)=0...
2026.05
96.2
ReasoningBank
label-flip ratio (ρ)=0...
2026.05
93.6
LiSA
label-flip ratio (ρ)=1...
2026.05
93.3
LiSA
label-flip ratio (ρ)=2...
2026.05
91.7
Synapse
label-flip ratio (ρ)=0...
2026.05
89.1
ReasoningBank
label-flip ratio (ρ)=1...
2026.05
88.2
Synapse
label-flip ratio (ρ)=1...
2026.05
87.8
Synapse
label-flip ratio (ρ)=2...
2026.05
86.3
AGrail
label-flip ratio (ρ)=0...
2026.05
85.7
ReasoningBank
label-flip ratio (ρ)=2...
2026.05
85.7
AGrail
label-flip ratio (ρ)=1...
2026.05
84.6
AGrail
label-flip ratio (ρ)=2...
2026.05
83.2
Pure Prediction
label-flip ratio (ρ)=0...
2026.05
81.8
Pure Prediction
label-flip ratio (ρ)=1...
2026.05
81.8
Pure Prediction
label-flip ratio (ρ)=2...
2026.05
81.8
Feedback
Search any
task
Search any
task