Beyond Predictive Uncertainty: Reliable Representation Learning with Structural Constraints

TL;DR

This paper advocates for treating the reliability of learned representations as a core property, introducing a framework that models representation uncertainty and uses structural constraints to enhance stability and robustness.

Contribution

It proposes a novel framework for reliable representation learning that incorporates uncertainty modeling and structural constraints as regularizers, independent of specific architectures.

Findings

Encourages stable, well-calibrated representations.

Reduces spurious variability through structural constraints.

Enhances robustness to noise and perturbations.

Abstract

Uncertainty estimation in machine learning has traditionally focused on the prediction stage, aiming to quantify confidence in model outputs while treating learned representations as deterministic and reliable by default. In this work, we challenge this implicit assumption and argue that reliability should be regarded as a first-class property of learned representations themselves. We propose a principled framework for reliable representation learning that explicitly models representation-level uncertainty and leverages structural constraints as inductive biases to regularize the space of feasible representations. Our approach introduces uncertainty-aware regularization directly in the representation space, encouraging representations that are not only predictive but also stable, well-calibrated, and robust to noise and structural perturbations. Structural constraints, such as sparsity, relational structure, or feature-group dependencies, are incorporated to define meaningful geometry and reduce spurious variability in learned representations, without assuming fully correct or noise-free structure. Importantly, the proposed framework is independent of specific model architectures and can be integrated with a wide range of representation learning methods.