Pith. sign in

REVIEW 5 cited by

I-Con: A Unifying Framework for Representation Learning

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2504.16929 v1 pith:FD5LAPVA submitted 2025-04-23 cs.LG cs.AIcs.CVcs.ITmath.IT

I-Con: A Unifying Framework for Representation Learning

classification cs.LG cs.AIcs.CVcs.ITmath.IT
keywords learningdifferentframeworkfunctionslossmethodsrepresentationclasses
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

As the field of representation learning grows, there has been a proliferation of different loss functions to solve different classes of problems. We introduce a single information-theoretic equation that generalizes a large collection of modern loss functions in machine learning. In particular, we introduce a framework that shows that several broad classes of machine learning methods are precisely minimizing an integrated KL divergence between two conditional distributions: the supervisory and learned representations. This viewpoint exposes a hidden information geometry underlying clustering, spectral methods, dimensionality reduction, contrastive learning, and supervised learning. This framework enables the development of new loss functions by combining successful techniques from across the literature. We not only present a wide array of proofs, connecting over 23 different approaches, but we also leverage these theoretical results to create state-of-the-art unsupervised image classifiers that achieve a +8% improvement over the prior state-of-the-art on unsupervised classification on ImageNet-1K. We also demonstrate that I-Con can be used to derive principled debiasing methods which improve contrastive representation learners.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Understanding Self-Supervised Learning via Latent Distribution Matching

    cs.LG 2026-05 unverdicted novelty 6.0

    Self-supervised learning is cast as latent distribution matching that aligns representations to a model while enforcing uniformity, unifying multiple SSL families and proving identifiability for predictive variants ev...

  2. Understanding Self-Supervised Learning via Latent Distribution Matching

    cs.LG 2026-05 unverdicted novelty 6.0

    Self-supervised learning is recast as latent distribution matching that unifies multiple SSL families and yields a sampling-free Kalman-based predictor plus an identifiability proof for predictive variants under mild ...

  3. Understanding Self-Supervised Learning via Latent Distribution Matching

    cs.LG 2026-05 conditional novelty 6.0

    Self-supervised learning can be understood as latent distribution matching, and under a Gaussian predictive model this yields identifiable representations up to affine transformations.

  4. Learning Minimal Representations of Fermionic Ground States

    quant-ph 2025-12 conditional novelty 6.0

    Autoencoders trained on Hubbard ground-state measurement vectors show a sharp reconstruction threshold at L−1 latent dimensions, and the decoder can be used as a variational ansatz for energy minimization.

  5. Understanding Self-Supervised Learning via Latent Distribution Matching

    cs.LG 2026-05 unverdicted novelty 5.0

    Self-supervised learning is recast as latent distribution matching that unifies ICA with contrastive and predictive methods and derives an identifiable nonlinear Bayesian filtering model for timeseries.