Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Science Question Answering on ARC Challenge (Accuracy)
Loading...
66.1
Accuracy (ARC)
No pruning
23.356
34.453
45.55
56.647
Mar 17, 2026
Mar 18, 2026
Mar 19, 2026
Mar 21, 2026
Mar 22, 2026
Mar 23, 2026
Mar 25, 2026
Accuracy (ARC)
Updated 23d ago
Evaluation Results
Method
Method
Links
Accuracy (ARC)
No pruning
Sparsity=0, Backbone=G...
2026.03
66.1
Fanar-27B
Size=27B, Few-shot set...
2026.03
65.61
Fanar-1-9b-instruct
Size=9B, Few-shot sett...
2026.03
65.19
Llama-3.3-70b-Instruct
Size=70B, Few-shot set...
2026.03
63.05
Qwen3-32B
Size=32B, Few-shot set...
2026.03
60.84
AceGPT-v2-70B-Chat
Size=70B, Few-shot set...
2026.03
60.07
Gemma-3-27B-it
Size=27B, Few-shot set...
2026.03
59.98
Jais-2-70B-Chat
Size=70B, Few-shot set...
2026.03
59.3
Allam-7B-Instruct-preview-v2
Size=7B, Few-shot sett...
2026.03
58.62
AceGPT-v2-32B-Chat
Size=32B, Few-shot set...
2026.03
53.92
Magnitude-Dim
Sparsity=10%, Backbone...
2026.03
50.6
DIET
Sparsity=10%, Backbone...
2026.03
49.6
Karnak
Size=40B, Few-shot set...
2026.03
47.35
PuDDing
Sparsity=20%, Backbone...
2026.03
32.1
DIET
Sparsity=20%, Backbone...
2026.03
32
Magnitude-Dim
Sparsity=20%, Backbone...
2026.03
30.8
PuDDing
Sparsity=10%, Backbone...
2026.03
30.3
SliceGPT
Sparsity=20%, Backbone...
2026.03
25
SliceGPT
Sparsity=10%, Backbone...
2026.03
25
Feedback
Search any
task
Search any
task