National Research University Higher School of Economics

For some computable patterns the per-bit errors of universal sequence prediction do not sum to a bounded value, yet their squares are bounded exponentially in the pattern's program length, and polynomially when a computable classification function fixes the bits.

For some computable patterns the per-bit errors of universal sequence prediction do not sum to a bounded value, yet their squares are bounded exponentially in the pattern's program length, and polynomially when a computable classification function fixes the bits.

A systematic re-evaluation across 41 tabular datasets under one protocol finds quantile loss gives more consistent test-RMSE gains over tuned mean squared error on MLP, TabPack, RealMLP and TabM, and reveals that regions with different noise levels prefer different stopping epochs.

A systematic re-evaluation across 41 tabular datasets under one protocol finds quantile loss gives more consistent test-RMSE gains over tuned mean squared error on MLP, TabPack, RealMLP and TabM, and reveals that regions with different noise levels prefer different stopping epochs.

The work introduces Tactile-JEPA, a self-supervised pretraining method for distributed tactile sensors (e-skins) that samples masks at local and global scales over the taxel connectivity graph and predicts the embeddings of masked taxels; across magnetic and piezoresistive sensors and three datasets (Sparsh-skin, Tactile socks, DECO-50) it reduces force-estimation RMSE by 6.3% and in-hand pose RMSE by 20.8% over the strongest prior baseline, with consistent gains in action classification, object classification, and tactile-conditioned policy learning.

The work introduces Tactile-JEPA, a self-supervised pretraining method for distributed tactile sensors (e-skins) that samples masks at local and global scales over the taxel connectivity graph and predicts the embeddings of masked taxels; across magnetic and piezoresistive sensors and three datasets (Sparsh-skin, Tactile socks, DECO-50) it reduces force-estimation RMSE by 6.3% and in-hand pose RMSE by 20.8% over the strongest prior baseline, with consistent gains in action classification, object classification, and tactile-conditioned policy learning.