Closing the Context Gap: Activation Alignment for Tabular In-Context Learning
Activation alignment lets tabular models use less context while recovering much of full-context accuracy.
2 stories, newest first · one card per story
Activation alignment lets tabular models use less context while recovering much of full-context accuracy.
TabFM, a 400M tabular foundation model, delivers calibrated zero-shot predictions that beat tuned AutoML.
Today 793 readers · 868 views · 84 reads · all time 215K views