Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Lifelong Knowledge Editing on GPT-J 1024 Edits (test)
Loading...
14.6
Score (S)
WilKE
-0.584
3.358
7.3
11.242
Feb 16, 2024
Score (S)
Effectiveness (ES)
Generality (PS)
Locality (NS)
Retention (ERS)
Retention (ORS)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Score (S)
Effectiveness (ES)
Generality (PS)
Locality (NS)
Retention (ERS)
Retention (ORS)
WilKE
Backbone=GPT-J (6B)
2024.02
14.6
71.3
50.6
7.4
19.1
8.5
ROME
Backbone=GPT-J (6B)
2024.02
8.7
28.7
20.7
4.6
10.9
5.8
MEND
Backbone=GPT-J (6B)
2024.02
3.7
3.7
1.7
4.7
9.8
9.2
MEMIT
Backbone=GPT-J (6B)
2024.02
0
32.1
23.8
9.3
0
0
Feedback
Search any
task
Search any
task
