Skip to content

Repository files navigation

PACENet

This repository includes the official implementation of our paper “PACEAttention: Principled and Adaptive Feature Compression-Expansion Grounded in the Geometry of $\text{MCR}^2$”[ICML 2026, paper link]

Inspired by the geometric interpretation of the gradients of the maximal coding rate reduction ($\text{MCR}^2$) objective, we propose a principled and interpretable attention mechanism that facilitates feature compression and expansion.

Lay Summary

Deep neural networks are often powerful but difficult to interpret, making it unclear why they work well or how their architectures should be designed. In this work, we propose PACEAttention, a new attention mechanism motivated by geometric principles from representation learning theory.

Our method entails feature learning as a progressive process, where same-category feature representations are compressed into compact subspaces, while cross-category representations are expanded. This process is guided by the intrinsic structure of the data. The resulting model, PACENet, is both principled and efficient, and produces competitive performance on several classification benchmarks. This design also makes PACENet easier to interpret, with different parts of the model naturally focusing on different image regions and fine details, even when trained using standard supervised learning.

One Layer of PACENet

df

About

This repository includes the official implementation our paper “PACEAttention: Principled and Adaptive Feature Compression-Expansion Grounded in the Geometry of $\text{MCR}^2$”

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages