Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Question Answering on SQuAD (2048 tokens)
Loading...
49.4
Accuracy
Looped Transformer
38.688
41.469
44.25
47.031
May 20, 2026
Accuracy
Updated 12d ago
Evaluation Results
Method
Method
Links
Accuracy
Looped Transformer
Model (1.5B)=1.5B, Loo...
2026.05
49.4
Looped Hybrid (Full+GDN)
Model (1.5B)=1.5B, Loo...
2026.05
48.9
Looped Hybrid (GDN+DSA)
Model (1.5B)=1.5B, Loo...
2026.05
48
Transformer
Model (1.5B)=1.5B, Loo...
2026.05
46.6
Looped GDN
Model (1.5B)=1.5B, Loo...
2026.05
41.8
Looped Mamba-2
Model (1.5B)=1.5B, Loo...
2026.05
40.5
GDN
Model (1.5B)=1.5B, Loo...
2026.05
40
Mamba-2
Model (1.5B)=1.5B, Loo...
2026.05
39.1
Feedback
Search any
task
Search any
task