Tydra: An Efficient Hybrid Model for Tabular Data
Transformer-based tabular foundation models such as TabPFN achieve strong predictive performance but incur quadratic computational cost with context length. On the other hand, subquadratic SSM-based alternatives such as Hydra trade away accuracy for efficiency. To balance both, we introduce Tydra, a hybrid Transformer-State Space Model (SSM) architecture for tabular in-context learning that interleaves attention and SSM layers. Across 30 OpenML datasets, Tydra reduces inference time by 30% relative to TabPFN while retaining much of its predictive performance. Tydra also outperforms an approxim
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 53%Google's TabFM skips per-dataset training and still predicts on tables it's never seen →
- PossiblePossibly related (embedding) · 52%TabFM Studio: point-and-click predictions on spreadsheets with tabular foundation models, fully local [P] →
- LinkedLinked via arxiv author · 85%Mieszko Komisarczyk →
“Tydra: An Efficient Hybrid Model for Tabular Data”
- LinkedLinked via arxiv author · 85%Saurabh Mathur →
“Tydra: An Efficient Hybrid Model for Tabular Data”
- LinkedLinked via arxiv author · 85%Maurice Kraus →
“Tydra: An Efficient Hybrid Model for Tabular Data”
- LinkedLinked via arxiv author · 85%Sriraam Natarajan →
“Tydra: An Efficient Hybrid Model for Tabular Data”
- LinkedLinked via arxiv author · 85%Kristian Kersting →
“Tydra: An Efficient Hybrid Model for Tabular Data”