Tabular foundation models use distinct similarity-based readouts such as attention-weighted votes or class-conditional means, with invariances tracing to removable positional parameters.
International Conference on Machine Learning , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Introduces effective dimension d_ρ from spectral analysis of reasoning trajectories to distinguish task hardness (0.93 AUC on MATH500) and uses kinematic features for early correctness prediction from partial generations.
citing papers explorer
-
A Mechanistic Study of Tabular Foundation Models
Tabular foundation models use distinct similarity-based readouts such as attention-weighted votes or class-conditional means, with invariances tracing to removable positional parameters.
-
Geometric Signatures of Reasoning: A Spectral Perspective on Task Hardness
Introduces effective dimension d_ρ from spectral analysis of reasoning trajectories to distinguish task hardness (0.93 AUC on MATH500) and uses kinematic features for early correctness prediction from partial generations.