Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Relational linear probing on LANG
Loading...
98
F1 (GT)
KL-RP
-2.88
23.31
49.5
75.69
May 21, 2026
F1 (GT)
F1 (LLM)
dKL
Updated 12d ago
Evaluation Results
Method
Method
Links
F1 (GT)
F1 (LLM)
dKL
KL-RP
Model=Llama-3.1, Layer=16
2026.05
98
98
0.06
KL-RP
Model=Gemma-2, Layer=13
2026.05
51
62
2.19
LRE
Model=Llama-3.1, Layer=16
2026.05
6
36
0.51
Random
Model=Llama-3.1, Layer=16
2026.05
4
5
2.04
Random
Model=Gemma-2, Layer=13
2026.05
3
13
6.16
LRE
Model=Gemma-2, Layer=13
2026.05
1
48
0.45
Feedback
Search any
task
Search any
task
