Causilo speeds up accurate predictions on tabular data
Causilo Technical Report
Machine Learning
Summary
Predicting outcomes from tables of data is important but can be slow or less accurate. The authors present Causilo, a new model that balances fast processing with strong prediction accuracy on many benchmark tests. It uses a special way to look at the data both by columns and rows, refining the information efficiently. Causilo was trained on millions of made-up tables and performs better and faster than comparable models.
What this means in practice
- •For data engineering teams: Speed up predictions in production systems that analyze structured tables without sacrificing accuracy.
- •For business intelligence developers: Build faster analytics tools that provide reliable forecasts from complex table datasets.
Authors
Minyong Cho, Minho Jeong, Dooho Lee, Jinmo Lee, Jaemin Yoo
Abstract
We introduce Causilo, a tabular foundation model (TFM) that combines frontier predictive performance with exceptionally fast inference. On TabArena, Causilo achieves 1785.4 Elo, at a median inference time of 0.10 seconds per 1K test samples. It outperforms TabPFN-3.5-Fast with 31.6% less inference time, placing it on the performance--efficiency Pareto frontier. Causilo follows TabICL's column-then-row architecture but introduces another row-refinement module before row compression. This module exchanges information among cell representations within each row after column encoding. The refined cells then visit the context set again through an additional column stage before being compressed into row embeddings. For inference efficiency, both row stages use cross-attention through a fixed number of summary tokens, keeping their attention cost linear in the number of features. Pretrained on approximately 36M synthetic tables, Causilo delivers strong benchmark results across TabArena, BeyondArena, and ScoringBench, achieving frontier-level performance with substantially faster inference.