Self-supervised learning (SSL) has emerged as a powerful approach for learning biologically meaningful representations of single-cell data. To establish best practices in this domain, we present a comprehensive benchmark evaluating eight SSL methods across three downstream tasks and eight datasets, with various data augmentation strategies. Our results demonstrate that SimCLR and VICReg consistently outperform other methods across different tasks. Furthermore, we identify random masking as the most effective augmentation technique. This benchmark provides valuable insights into the application of SSL to single-cell data analysis, bridging the gap between SSL and single-cell biology.

Philip Toma^*, Olga Ovcharenko^*, Imant Daunhawer, Julia Vogt, Florian Barkmann^†, Valentina Boeva^†
* denotes shared first authorship, ^† denotes shared last authorship

Preprint

06.11.2024

DOICode

Supervised learning has become a cornerstone of modern machine learning, yet a comprehensive theory explaining its effectiveness remains elusive. Empirical phenomena, such as neural analogy-making and the linear representation hypothesis, suggest that supervised models can learn interpretable factors of variation in a linear fashion. Recent advances in self-supervised learning, particularly nonlinear Independent Component Analysis, have shown that these methods can recover latent structures by inverting the data generating process. We extend these identifiability results to parametric instance discrimination, then show how insights transfer to the ubiquitous setting of supervised learning with cross-entropy minimization. We prove that even in standard classification tasks, models learn representations of ground-truth factors of variation up to a linear transformation. We corroborate our theoretical contribution with a series of empirical studies. First, using simulated data matching our theoretical assumptions, we demonstrate successful disentanglement of latent factors. Second, we show that on DisLib, a widely-used disentanglement benchmark, simple classification tasks recover latent structures up to linear transformations. Finally, we reveal that models trained on ImageNet encode representations that permit linear decoding of proxy factors of variation. Together, our theoretical findings and experiments offer a compelling explanation for recent observations of linear representations, such as superposition in neural networks. This work takes a significant step toward a cohesive theory that accounts for the unreasonable effectiveness of supervised deep learning.

Patrik Reizinger^*, Alice Bizeul^*, Attila Juhos^*, Julia E. Vogt, Randall Balestriero, Wieland Brendel, David Klindt
^* denotes shared first authorship

04.11.2024

Link

Ričards Marcinkevičs^*, Kacper Sokol^*, Akhil Paulraj, Melinda A. Hilbert, Vivien Rimili, Sven Wellmann, Christian Knorr, Bertram Reingruber, Julia E. Vogt^†, Patricia Reis Wolfertstetter^†
* denotes shared first authorship, ^† denotes shared last authorship

medRxiv

29.10.2024

LinkDOI

Self-Supervised Learning (SSL) methods often consist of elaborate pipelines with hand-crafted data augmentations and computational tricks. However, it is unclear what is the provably minimal set of building blocks that ensures good downstream performance. The recently proposed instance discrimination method, coined DIET, stripped down the SSL pipeline and demonstrated how a simple SSL algorithm can work by predicting the sample index. Our work proves that DIET recovers cluster-based latent representations, while successfully identifying the correct cluster centroids in its classification head. We demonstrate the identifiability of DIET on synthetic data adhering to and violating our assumptions, revealing that the recovery of the cluster centroids is even more robust than the feature recovery.

Attila Juhos^*, Alice Bizeul^*, Patrik Reizinger^*, David Klindt, Randall Balestriero, Mark Ibrahim, Julia E Vogt, Wieland Brendel
^* denotes shared first authorship

NeurIPS 2024 Workshop: Self-Supervised Learning-Theory and Practice

10.10.2024

Link

The structure of many real-world datasets is intrinsically hierarchical, making the modeling of such hierarchies a critical objective in both unsupervised and supervised machine learning. Recently, novel approaches for hierarchical clustering with deep architectures have been proposed. In this work, we take a critical perspective on this line of research and demonstrate that many approaches exhibit major limitations when applied to realistic datasets, partly due to their high computational complexity. In particular, we show that a lightweight procedure implemented on top of pre-trained non-hierarchical clustering models outperforms models designed specifically for hierarchical clustering. Our proposed approach is computationally efficient and applicable to any pre-trained clustering model that outputs logits, without requiring any fine-tuning. To highlight the generality of our findings, we illustrate how our method can also be applied in a supervised setup, recovering meaningful hierarchies from a pre-trained ImageNet classifier.

Emanuele Palumbo, Moritz Vandenhirtz, Alain Ryser, Imant Daunhawer^†, Julia E. Vogt^†
† denotes shared last authorship

Preprint

10.10.2024

DOI