Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
LLM Unlearning on TOFU (Metrics: Agg., Mem., Priv., Util.)
Loading...
61
Aggregated Score
Retain
-2.44
14.03
30.5
46.97
May 7, 2026
Aggregated Score
Memory Score
Privacy Score
Utility Score
Updated 26d ago
Evaluation Results
Method
Method
Links
Aggregated Score
Memory Score
Privacy Score
Utility Score
Retain
Backbone=Llama-3.2-1B-...
2026.05
61
35
100
99
ROSU
Backbone=Llama-3.2-1B-...
2026.05
58
38
74
82
RMU
Backbone=Llama-3.2-1B-...
2026.05
54
48
53
62
UAM
Backbone=Llama-3.2-1B-...
2026.05
50
38
62
57
SimNPO
Backbone=Llama-3.2-1B-...
2026.05
49
36
44
99
UNDIAL
Backbone=Llama-3.2-1B-...
2026.05
44
30
46
79
IdkNLL
Backbone=Llama-3.2-1B-...
2026.05
13
10
8
91
PCGrad
Backbone=Llama-3.2-1B-...
2026.05
6
98
2
61
AltPO
Backbone=Llama-3.2-1B-...
2026.05
5
67
2
94
IdkDPO
Backbone=Llama-3.2-1B-...
2026.05
3
56
1
93
NPO
Backbone=Llama-3.2-1B-...
2026.05
2
52
1
97
Init. ft.
Backbone=Llama-3.2-1B-...
2026.05
0
0
1
100
GradDiff
Backbone=Llama-3.2-1B-...
2026.05
0
99
0
77
Feedback
Search any
task
Search any
task
