aiqbench analysis
ProtoSSL introduces a self-supervised framework for training projection-based prototype models, significantly improving label efficiency and interpretability in low-data regimes. By separating motif discovery from task-specific alignment, the method allows foundational prototype banks to be pretrained on unlabeled data before fine-tuning for specific applications like ECG analysis. This approach advances the field of machine learning by enabling more data-efficient and explainable deep neural networks without relying on extensive labeled datasets.
ProtoSSL: Self-Supervised Pretraining and Downstream Transfer for Projection-Based Prototype Models
arXiv:2605.06943v2 Announce Type: replace Abstract: In domains where both predictive performance and interpretability are essential, deep neural networks achieve strong results but provide limited insight into how their predictions are made. Projection-based prototype networks address this limitation by grounding predictions in similarity to representative training examples, enabling case-based explanations and global prototype inspection. However, existing approaches rely on label supervision, tying prototypes to a specific task and requiring large labeled datasets. We introduce ProtoSSL, a framework for pretraining a foundational latent prototype bank on unlabeled data and transferring it to downstream tasks to create interpretable, projection-based prototype models. Our key idea is to separate motif discovery from label alignment. ProtoSSL first learns a transferrable prototype bank using a self-supervised objective applied directly to prototype activations, and then aligns these prototypes to downstream tasks through a novel efficient assignment procedure. Across six electrocardiography (ECG) datasets, ProtoSSL improves label efficiency, outperforming supervised prototype baselines in low-data regimes with as few as 256 labeled examples; with fine-tuning, ProtoSSL outperforms supervised prototype baselines at full dataset scale. In a human evaluation study, ProtoSSL produces prototypes and prototype-based explanations that are judged more favorably than those learned with direct label supervision. We further show that the framework extends to audio classification. Thus, ProtoSSL enables both learning foundational prototypes from unlabeled data before the downstream label space is known, and subsequent assignment to new tasks to create interpretable, projection-grounded prototypes.
Abstract:In domains where both predictive performance and interpretability are essential, deep neural networks achieve strong results but provide limited insight into how their predictions are made. Projection-based prototype networks address this limitation by grounding predictions in similarity to representative training examples, enabling case-based explanations and global prototype inspection. However, existing approaches rely on label supervision, tying prototypes to a specific task and requiring large labeled datasets. We introduce ProtoSSL, a framework for pretraining a foundational latent prototype bank on unlabeled data and transferring it to downstream tasks to create interpretable, projection-based prototype models. Our key idea is to separate motif discovery from label alignment. ProtoSSL first learns a transferrable prototype bank using a self-supervised objective applied directly to prototype activations, and then aligns these prototypes to downstream tasks through a novel efficient assignment procedure. Across six electrocardiography (ECG) datasets, ProtoSSL improves label efficiency, outperforming supervised prototype baselines in low-data regimes with as few as 256 labeled examples; with fine-tuning, ProtoSSL outperforms supervised prototype baselines at full dataset scale. In a human evaluation study, ProtoSSL produces prototypes and prototype-based explanations that are judged more favorably than those learned with direct label supervision. We further show that the framework extends to audio classification. Thus, ProtoSSL enables both learning foundational prototypes from unlabeled data before the downstream label space is known, and subsequent assignment to new tasks to create interpretable, projection-grounded prototypes.
Machine Learning (cs.LG)
arXiv:2605.06943 [cs.LG]
(or arXiv:2605.06943v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2605.06943 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 321 words
Related coverage
Neonatal Hypoxic-ischaemic Encephalopathy Classification from the EEG and HRV Signals Using a Conformer based Masked Autoencoder
arXiv · Jul 28, 2026
BrainNext: A General-Purpose Self-Supervised Foundation Model for Brain MRI Analysis
arXiv · Jul 21, 2026
MSBraM: A Multi-scale Self-supervised Brain Foundation Model for Hierarchical EEG Dynamics Learning
arXiv · Jul 24, 2026
A Light Bilevel Refinement Aligns Self-Supervised Representations for Stronger Task-Specific Learning
arXiv · Sep 30, 2026