KnowsTFM: Knowledge-Informed Fine-Tuning of Small Tabular Foundation Models
Tabular foundation models have advanced deep learning for tabular data by delivering strong default performance across many small and medium tasks. Yet in niche domains, where data is scarce, high-dimensional, and shifted from the pretraining distribution, they may still fail to outperform carefully designed domain-specific methods. Many such domains also provide curated relational knowledge in the form of knowledge graphs and knowledge banks, but how to use this knowledge to improve and steer \textit{small} specialist tabular foundation models remains unclear. We address this problem through
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.
- LinkedLinked via unknownFine-tuning →
- PossiblePossibly related (embedding) · 29%PriorLabs/TabPFN →
“Possibly related via embedding similarity 0.57 (not asserted). Timestamp check: artifact after paper (+3d).”
- PossiblePossibly related (embedding) · 45%TabFM and the Rise of Tabular Foundation Models | by Adnan Masood, PhD. | Jul, 2026 - Medium →
- PossiblePossibly related (embedding) · 51%deepset-ai/haystack-core-integrations →
- PossiblePossibly related (embedding) · 58%Google's TabFM skips per-dataset training and still predicts on tables it's never seen →
- PossiblePossibly related (embedding) · 57%PriorLabs/tabpfn-extensions →
- PossiblePossibly related (embedding) · 47%TabFM Studio: point-and-click predictions on spreadsheets with tabular foundation models, fully local [P] →