Knowledge-style Retrieval on FDA 2048 tokens
62AccuracyLooped Hybrid (Full+GDN)
Evaluation Results
| Method | Links | |
|---|---|---|
| Looped Hybrid (Full+GDN)Model (1.5B)=1.5B, Looped=true, Iterations=T=4, Mixer=Hybrid2026.05 | 62 | |
| Looped TransformerModel (1.5B)=1.5B, Looped=true, Iterations=T=42026.05 | 61.7 | |
| Looped Hybrid (GDN+DSA)Model (1.5B)=1.5B, Looped=true, Iterations=T=4, Mixer=Hybrid2026.05 | 60.4 | |
| TransformerModel (1.5B)=1.5B, Looped=false2026.05 | 58.4 | |
| Looped GDNModel (1.5B)=1.5B, Looped=true, Iterations=T=42026.05 | 30.6 | |
| GDNModel (1.5B)=1.5B, Looped=false2026.05 | 28.3 | |
| Looped Mamba-2Model (1.5B)=1.5B, Looped=true, Iterations=T=42026.05 | 25.8 | |
| Mamba-2Model (1.5B)=1.5B, Looped=false2026.05 | 23.7 |