Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
LLM Unlearning on MT-Bench
Loading...
5.62
Fluency
Base Model
0.8152
2.0626
3.31
4.5574
Feb 19, 2025
Fluency
Updated 1mo ago
Evaluation Results
Method
Method
Links
Fluency
Base Model
Backbone=Llama-3-8B
2025.02
5.62
RMU
Backbone=Llama-3-8B
2025.02
5.39
NPOGDR
Backbone=Llama-3-8B
2025.02
5.18
GAGDR
Backbone=Llama-3-8B
2025.02
3.97
Base Model
Backbone=Mistral-7B
2025.02
1.71
RMU
Backbone=Mistral-7B
2025.02
1.58
NPOGDR
Backbone=Mistral-7B
2025.02
1.04
NPOKLR
Backbone=Llama-3-8B
2025.02
1.03
GAKLR
Backbone=Llama-3-8B
2025.02
1.01
TV
Backbone=Llama-3-8B
2025.02
1.01
GA
Backbone=Llama-3-8B
2025.02
1
NPO
Backbone=Llama-3-8B
2025.02
1
GA
Backbone=Mistral-7B
2025.02
1
GAGDR
Backbone=Mistral-7B
2025.02
1
GAKLR
Backbone=Mistral-7B
2025.02
1
NPO
Backbone=Mistral-7B
2025.02
1
NPOKLR
Backbone=Mistral-7B
2025.02
1
TV
Backbone=Mistral-7B
2025.02
1
Feedback
Search any
task
Search any
task
