Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Domain-specific Question Answering on QASC (Accuracy, Relative Accuracy Improvement (%))
Loading...
68.36
Accuracy
Vanilla Fine-tuning
-0.6544
17.2628
35.18
53.0972
May 26, 2025
Accuracy
Relative Accuracy Improvement
Updated 2mo ago
Evaluation Results
Method
Method
Links
Accuracy
Relative Accuracy Improvement
Vanilla Fine-tuning
Model=Llama3-8B, Unlea...
2025.05
68.36
-
Llama3.2-1B
Forgetting Task=none
2025.05
42.98
-
LWF
Forgetting Task=mixed
2025.05
5.54
-
LWF
Forgetting Task=dental
2025.05
5.28
-
LWF
Forgetting Task=gsm8k
2025.05
4.03
-
LWF
Forgetting Task=sst5
2025.05
3.02
-
LWF
Forgetting Task=psychol
2025.05
2
-
LWF
Model=Llama3-8B, Unlea...
2025.05
-
5.37
LWF
Model=Llama3-8B, Unlea...
2025.05
-
2.68
LWF
Model=Llama3-8B, Unlea...
2025.05
-
1.26
LWF
Model=Llama3-8B, Unlea...
2025.05
-
4.42
LWF
Model=Llama3-8B, Unlea...
2025.05
-
7.9
Feedback
Search any
task
Search any
task
