Multiple Instance Learning (MIL) has emerged as the best solution for Whole Slide Image (WSI) classification. It consists of dividing each slide into patches, which are treated as a bag of instances labeled with a global label. MIL includes two main approaches: instance-based and embedding-based. In the former, each patch is classified independently, and then the patch scores are aggregated to predict the bag label. In the latter, bag classification is performed after aggregating patch embeddings. Even if instance-based methods are naturally more interpretable, embedding-based MILs have usually been preferred in the past due to their robustness to poor feature extractors. However, recently, the quality of feature embeddings has drastically increased using self-supervised learning (SSL). Nevertheless, many authors continue to endorse the superiority of embedding-based MIL. To investigate this further, we conduct 710 experiments across 4 datasets, comparing 10 MIL strategies, 6 self-supervised methods with 4 backbones, 4 foundation models, and various pathology-adapted techniques. Furthermore, we introduce 4 instance-based MIL methods never used before in the pathology domain. Through these extensive experiments, we show that with a good SSL feature extractor, simple instance-based MILs, with very few parameters, obtain similar or better performance than complex, state-of-the-art (SOTA) embedding-based MIL methods, setting new SOTA results on the BRACS and Camelyon16 datasets. Since simple instance-based MIL methods are naturally more interpretable and explainable to clinicians, our results suggest that more effort should be put into well-adapted SSL methods for WSI rather than into complex embedding-based MIL methods. Keywords: Whole Slide Image Classification, Self-Supervised Learning, Multiple Instance Learning, Digital Pathology 1 Introduction Whole Slide histopathology Image (WSI) analysis has become an increasingly common tool for disease diagnosis in digital pathology [1] . However, the gigapixel size of WSIs, makes the manual analysis very time-consuming and presents significant challenges for conventional Deep Learning (DL) methods [2, 3], as they are not designed to support such large images. To address that, a simple approach involves dividing the WSI into smaller patches that DL methods can easily handle. Then, features or predictions from a patch-level encoder/classifier are aggregated to get the slide-level prediction [4, 5]. Nonetheless, this method requires very expensive patch-level annotations, which are not always available. Please note that a naive assignment of the slide label to all patches might be clinically incorrect, since the tissue section characterizing a disease might only occupy a small fraction of the slide, while all other patches should be considered as healthy. 1 arXiv:2505.01109v1

As foundation models gain prominence, Federated Foundation Models (FedFM) have emerged as a privacy-preserving approach to collaboratively fine-tune models in federated learning (FL) frameworks using distributed datasets across clients. A key challenge for FedFM, given the versatile nature of foundation models, is addressing out-of-distribution (OOD) generalization, where unseen tasks or clients may exhibit distribution shifts leading to suboptimal performance. Although numerous studies have explored OOD generalization in conventional FL, these methods are inadequate for FedFM due to the challenges posed by large parameter scales and increased data heterogeneity. To address these, we propose FedOA, which employs adapter-based parameter-efficient fine-tuning methods for efficacy and introduces personalized adapters with feature distance-based regularization to align distributions and guarantee OOD generalization for each client. Theoretically, we demonstrate that the conventional aggregated global model in FedFM inherently retains OOD generalization capabilities, and our proposed method enhances the personalized model's OOD generalization through regularization informed by the global model, with proven convergence under general non-convex settings. Empirically, the effectiveness of the proposed method is validated on benchmark datasets across various NLP tasks.

User sequence modeling is crucial for modern large-scale recommendation systems, as it enables the extraction of informative representations of users and items from their historical interactions. These user representations are widely used for a variety of downstream tasks to enhance users' online experience. A key challenge for learning these representations is the lack of labeled training data. While self-supervised learning (SSL) methods have emerged as a promising solution for learning representations from unlabeled data, many existing approaches rely on extensive negative sampling, which can be computationally expensive and may not always be feasible in real-world scenario. In this work, we propose an adaptation of Barlow Twins, a state-of-the-art SSL methods, to user sequence modeling by incorporating suitable augmentation methods. Our approach aims to mitigate the need for large negative sample batches, enabling effective representation learning with smaller batch sizes and limited labeled data. We evaluate our method on the MovieLens-1M, MovieLens-20M, and Yelp datasets, demonstrating that our method consistently outperforms the widely-used dual encoder model across three downstream tasks, achieving an 8%-20% improvement in accuracy. Our findings underscore the effectiveness of our approach in extracting valuable sequence-level information for user modeling, particularly in scenarios where labeled data is scarce and negative examples are limited.

Recent work has shown that successful per-domain generalizing action policies can be learned. Scaling behavior, from small training instances to large test instances, is the key objective; and the use of validation instances larger than training instances is one key to achieve it. Prior work has used fixed validation sets. Here, we introduce a method generating the validation set dynamically, on the fly, increasing instance size so long as informative and feasible. We also introduce refined methodology for evaluating scaling behavior, generating test instances systematically to guarantee a given confidence in coverage performance for each instance size. In experiments, dynamic validation improves scaling behavior of GNN policies in all 9 domains used.

Contrastive methods, dimensionality reduction algorithms like t-SNE, and clustering objectives such as k-Means all implicitly or explicitly define distributions over neighborhoods and minimize some divergence between them. The Information Contrastive Learning (I-Con) framework recently unified many such approaches by expressing them as the minimization of KL divergence between a fixed supervisory distribution p (j | i) and a learned distribution q ( j |i) over data neighborhoods [1]. However, I-Con treats this KL alignment statically, as if each point-wise loss were independent. In this paper, we propose a deeper view: that representation learning is fundamentally a process of recursive divergence minimization across a structured field of conditional distributions. Each neighborhood distribution depends on prior learned representations, forming a dynamic system that we call Recursive KL Divergence Optimization (RKDO). While the exponential moving average (EMA) recursion we employ has been used in several well-known self-supervised and semi-supervised methods such as Temporal Ensembling [2], Mean Teacher [3], and momentum-based frameworks like MoCo [4], BYOL [5], and DINO [6], our novel contribution lies in applying this recursive structure to the entire response field (the joint conditional distribution over representation pairs), rather than to individual weights or per-sample predictions. RKDO captures the temporal dynamics of representation learning that are absent in static frameworks, with significant implications for optimization efficiency. Our contributions include: A new theoretical framework that generalizes representation learning as recursive alignment of conditional distributions across the entire response field Mathematical formulations showing how RKDO captures temporal dynamics absent in static frameworks, with a formal proof of linear-rate convergence under this recursion Empirical evidence that RKDO's recursive approach results in significantly lower loss values (approximately 30% reduction across all tested datasets) Demonstration that RKDO requires 60-80% fewer computational resources (training epochs) to achieve results comparable to longer I-Con training Analysis of the trade-offs between optimization efficiency and generalization in recursive versus static approaches Our experiments suggest that while I-Con effectively represents a unified view of many typical representation learning approaches, RKDO can provide substantial efficiency improvements: achieving comparable optimization objectives with approximately 30% lower loss values, while potentially reducing computational requirements by 60-80% in the specific scenarios we studied. 2 Background and Related Work The KL divergence [7] is a foundational object in representation learning.

We present Variational Self-Supervised Learning (VSSL), a novel framework that combines variational inference with self-supervised learning to enable efficient, decoder-free representation learning. Unlike traditional VAEs that rely on input reconstruction via a decoder, VSSL symmetrically couples two encoders with Gaussian outputs. A momentum-updated teacher network defines a dynamic, data-dependent prior, while the student encoder produces an approximate posterior from augmented views. The reconstruction term in the ELBO is replaced with a cross-view denoising objective, preserving the analytical tractability of Gaussian KL divergence. We further introduce cosine-based formulations of KL and log-likelihood terms to enhance semantic alignment in high-dimensional latent spaces. Experiments on CIFAR-10, CIFAR-100, and ImageNet-100 show that VSSL achieves competitive or superior performance to leading self-supervised methods, including BYOL and MoCo V3. VSSL offers a scalable, probabilistically grounded approach to learning transferable representations without generative reconstruction, bridging the gap between variational modeling and modern self-supervised techniques.

Chronic obstructive pulmonary disease (COPD) is a major global health concern, with accurate severity assessment crucial for effective management, especially in intensive care units (ICUs). This study presents a novel approach to COPD sever - ity classification using machine learning algorithms applied to the MIMIC - III dataset. Our work presents a new application of the MIMIC - III dataset and con - tributes to the growing field of artificial intelligence in critical care medicine. We developed a model to classify COPD severity based on available ICU parameters, including blood gas measurements and vital signs. Our methodology incorpo - rated semi - supervised learning techniques to leverage unlabeled data, enhancing model robustness. A random forest classifier demonstrated superior performance, achieving 92.51% accuracy and 0.98 ROC AUC distinguishing between mild - to - moderate and severe COPD cases. This approach offers a practical, accurate, and accessible tool for rapid COPD severity assessment in ICU settings, poten - tially improving clinical decision - making and patient outcomes. Future research should focus on external validation and integration into clinical decision support systems to enhance COPD management in the ICUs.

Speaker diarization, a core problem in speech processing, entails partitioning a given audio stream according to the speakers. Even though progress has been made in the development of the models for high - resource languages, there is still a set of specific difficulties in going through a similar process for low - resource languages such as Kurdish: there are very few annotated datasets available; the language has dialects; speakers use code - switching a lot. These challenges are met in this study by training the Wav2V ec 2.0 SSL model on a Ku rdish dataset prepared for this purpose. Thanks to transfer learning, it was possible to transfer multiling ual representations learnt in other languages to the phonetic and acoustic features of Kurdish speech. The general Diarization Error Rate (DER) was reduced by 7.2%, and the cluster purity increased by 13% when compared to the baseline algorithm. They show that making improvements in any state - of - the - art model can help in enhancing the performance of under - resourced languages. Implications of this work include transcription services for Kurdish - language media programs, as well as speaker segmentation in multilingual call centers, teleconferencing, and videoconferencing systems. Therefore, this work demonstrates that self - supervised and transfer techniques can improve speaker diarization for Kurdish and other low - resource languages with diverse features. The approach provides a ba se for building effective diarization systems in other understudied languages, which remai ns essential for speech technology's equity.

Self-supervised learning (SSL) has enabled the development of vision foundation models for Earth Observation (EO), demonstrating strong transferability across diverse remote sensing tasks. While prior work has focused on network architectures and training strategies, the role of dataset curation, especially in balancing and diversifying pre-training datasets, remains underexplored. In EO, this challenge is amplified by the redundancy and heavy-tailed distributions common in satellite imagery, which can lead to biased representations and inefficient training. In this work, we propose a dynamic dataset pruning strategy designed to improve SSL pre-training by maximizing dataset diversity and balance. Our method iteratively refines the training set without requiring a pre-existing feature extractor, making it well-suited for domains where curated datasets are limited or unavailable. We demonstrate our approach on the Sentinel-1 Wave Mode (WV) Synthetic Aperture Radar (SAR) archive, a challenging dataset dominated by ocean observations. We train models from scratch on the entire Sentinel-1 WV archive spanning 10 years. Across three downstream tasks, our results show that dynamic pruning improves both computational efficiency and representation quality, leading to stronger transferability. We also release the weights of OceanSAR-1, the first model in the OceanSAR family, a series of foundation models for ocean observation and analysis using SAR imagery, at github.com/galeio-research/OceanSAR-models/.

Sign language is the primary communication language for people with disabling hearing loss. Sign language recognition (SLR) systems aim to recognize sign gestures and translate them into spoken language. One of the main challenges in SLR is the scarcity of annotated datasets. T o address this issue, we propose a semi-supervised learning (SSL) approach for SLR (SSLR), employing a pseudo-label method to annotate unlabeled samples. The sign gestures are represented using pose information that encodes the signer's skeletal joint points. This information is used as input for the Transformer backbone model utilized in the proposed approach. T o demonstrate the learning capabilities of SSL across various labeled data sizes, several experiments were conducted using different percentages of labeled data with varying numbers of classes. The performance of the SSL approach was compared with a fully supervised learning-based model on the WLASL-100 dataset. The obtained results of the SSL model outperformed the supervised learning-based model with less labeled data in many cases.