Overview

Representation learning studies how to turn raw data—such as images, text, graphs, time series, and sensor signals—into features that make important patterns easier to use. Common approaches include contrastive and masked prediction, autoencoders, predictive latent models, sparse coding, and methods that impose structure such as geometry, hierarchy, or disentangled subspaces. Good representations can reduce label needs and support transfer across tasks, domains, and modalities, with evaluations spanning linear probes, retrieval, classification, and downstream control. A central tension is that compression and invariance can discard information a later task needs, so no single objective or representation geometry works best across settings.

Latest papers

855 papers91 papers in the last four weeks, down 14% on the four weeks before. 0.6% of all new papers.

Oct 31, 2026

Lightweight Representation Learning for Low Tucker Rank Tensors

Low Tucker rank tensors provide compact representations of multidimensional dependencies, but standard learning pipelines can incur substantial computational and memory overhead from redundant intermediate quantities. We reduce this overhead without modifying the underlying representation objective or parameterization. Our approach combines three complementary components: (i) sequential extraction of only task-required spectral components, avoiding unnecessary full decompositions; (ii) factor-wise multilinear contractions without materializing large Kronecker operators; and (iii) compact latent-state updates without storing full ambient tensor iterates. These components preserve the algebraic operations and learning trajectories of the original algorithms while reducing their computational and memory footprints. Across five low Tucker rank tensor estimators, our implementations match the original trajectories to numerical precision (relative discrepancies of 10−1610^{-16} to 10−1310^{-13}), while reducing per-iteration runtime by up to 3.12×3.12\times and resident memory by up to 616×616\times (99.8%99.8\%) on the main order-3 benchmark at n=100n=100, with memory reductions exceeding 900×900\times across the full evaluation suite. These results show that substantial computational redundancy can be eliminated from Tucker representation learning without changing the underlying model or learned representation.
Oct 31, 2026

ISBC: Interpretable Sparse Binary Coding of Pretrained Text Embeddings

Pretrained text embeddings encode semantic structure in high-dimensional continuous spaces whose coordinates are difficult to interpret. We introduce Interpretable Sparse Binary Coding (ISBC), which learns a small set of binary dimensions from frozen embeddings without taxonomy labels or downstream outcomes. A trainable encoder and affine decoder combine reconstruction with controls on activity, factor utilization, dependence, and activation uncertainty. Individual bits and their conjunctions select overlapping document groups; complete activation profiles provide finer contexts for conditional comparisons. We evaluate these representations across corpora, backbones, and code configurations through reconstruction, geometry, predictive probes, and document inspection. Empirical cases identify both distinctions within existing categories and shared content across their boundaries. In a mixed-domain corpus, the most frequent profiles predominantly identify source collections, while selected within-source comparisons reveal sentiment and migration contrasts. Selected associations persist in additional documents encoded without refitting. In an engagement application, bits alone improve prediction of log-transformed outcomes over a constant reference, while their combinations identify within-event content contrasts associated with engagement. The results support using binary coordinates and their combinations for empirical interpretation, with complete profiles providing a fine, partly initialization-dependent description rather than a unique discrete taxonomy.
Oct 31, 2026

Representative Learning for Distributed Support Vector Machines

Distributed support vector machine (SVM) learning is challenging under heterogeneous local data and limited communication. We develop a Representative Learning framework based on Lifted Empirical Risk Minimization, replacing each covariate by a local Gaussian distribution. For linear SVMs, the resulting expected hinge loss has an exact closed form, is smooth for nondegenerate lifting covariances, and converges to the ordinary hinge loss as the covariance vanishes. The lifted formulation enables clouded representatives that encode both location and covariance. We study a model-independent Mean Representative (MR) and a Score-Matching Representative (SMR) that matches the class-wise lifted gradient at its construction parameter. We show that SMR covariance scaling is stable with respect to within-group heterogeneity and that the gradient discrepancy between full data and frozen representatives is controlled by the maximum within-group diameter. Under strong convexity and smoothness, this yields an inexact-gradient convergence guarantee. Experiments under heterogeneous class allocation show that lifted representatives remain competitive with centralized solutions, while SMR more closely recovers the centralized lifted-SVM optimizer. We further extend the framework to nonlinear SVMs through a shared kernel dictionary, with experiments on Ionosphere and Australian Credit Approval showing analogous behavior. These results provide a distributional representative framework for communication-efficient distributed SVM learning under heterogeneity.
Oct 30, 2026

KINDER: Kernel-based Independence for Fair Representation Learning via Prototype-space Erasure

Concept erasure is a prominent approach to achieving fairness in machine learning through the removal of sensitive attributes from learned representations. Prior concept erasure methods typically define debiasing indirectly through the failure of a chosen adversary, probe, or fairness penalty, which makes the target of erasure dependent on a particular decoder family or optimization setup. We introduce KINDER, which directly removes sensitive-attribute information from intermediate representations with respect to a chosen family of prototype functionals, and is compatible with any deep model containing an intermediate feature space, independent of the architecture, task type, or training paradigm. KINDER operates in a random Fourier feature space, where it estimates a sensitive subspace from protected group prototypes and projects representations onto its orthogonal complement. Thus, it forms an explicit representation cleaning mechanism distinct from prior fairness control studies. We conduct extensive experiments across diverse settings and show the effectiveness of KINDER on (i) supervised and self-supervised settings, (ii) classification, regression, and image segmentation tasks, and (iii) diverse data modalities, including visual, textual, and tabular data. Code and trained models are available at: https://github.com/abtinmU/KINDER
Oct 30, 2026

ID2Pattern: Pretraining Generalizable Identity Behavior Patterns for Transaction Risk Foundation Models

Large-scale behavioral pretraining offers a promising route to generalizable representations in online payment systems. However, high-cardinality identity fields such as card, device, and account IDs, though rich in behavioral signals, remain underexploited. Their continually growing vocabularies and limited semantic meaning make existing vocabulary-based and semantics-based encodings ill-suited. Their generalizable signal lies not in specific values, but in the identity behavior patterns they exhibit across a transaction sequence. Building on this insight, we propose ID2Pattern, a pretraining framework that brings these patterns into transaction foundation models. ID2Pattern first encodes identity fields with a Local Anonymous ID (LAID) tokenizer, which captures how each identity appears and recurs within the sequence, independent of its specific value. It then pretrains a multi-scale pattern codebook on unlabeled sequences, which abstracts LAID sequences into discrete behavior patterns for downstream prediction. Experiments on an industrial dataset with over 100M sequences and six public datasets show that ID2Pattern consistently outperforms strong baselines, improving industrial by 4.13% and public datasets by 8.83% on average. Notably, ID2Pattern remains equally effective even when all identity values are unseen during training, highlighting its potential to generalize to new users and markets.
Oct 30, 2026

Binary Sparse Codes: Where are we on Energy Efficient Representations?

Biological neural systems achieve impressive computational capabilities under tight metabolic constraints, relying on sparse cortical representations that minimise total neural activity. This combination of efficiency and representational power has motivated growing interest in sparse coding as a computational strategy for artificial systems. Artificial models designed to produce binary sparse codes span three broad paradigms, theoretical methods grounded in efficient coding, deep learning methods trained via backpropagation, and biologically plausible methods that learn through local Hebbian rules, each yielding codes whose quality depends not only on achieving low activation density but also on preserving the informational structure of the input. The latter property, however, lacks an agreed-upon quantitative definition, leaving the field without a principled basis for comparing methods. We propose a set of metrics that formalise information preservation as similarity preservation, measuring how faithfully a binary sparse code reproduces the relational geometry of the input space, and apply this framework to compare four representative methods under matched conditions. The results reveal a consistent trade-off in which models that achieve high information preservation do so at the cost of sparsity, while models approaching the uniformity of an optimal random sparse code tend to sacrifice it. Competitive Sparse Codes, a biologically plausible approach, is the only architecture to occupy a region of near-maximal sparsity while retaining meaningful information preservation. These findings frame the simultaneous satisfaction of both factors as an open challenge for the field and provide a reusable evaluation framework for measuring progress toward it.
Oct 29, 2026

Grounding Predictive Reliability through Masked Observability

An uncertainty estimate is meaningful only if its semantics can be tied to observable predictive behavior. Normal-Inverse-Gamma (NIG) evidential regression provides a structured decomposition of predictive variability into data-scale and predictive-mean components, but marginal likelihood does not identify their allocation. We address this semantic ambiguity in masked representation learning. MORSE (Masked Observability for Reliable Structured Encoding) predicts the same hidden target under controlled changes in visible context and uses normalized cross-context variation to ground predictive-mean uncertainty, while a Gaussian objective learns the conditional mean and data scale. The normalized variation plays the role of the NIG factor 1/ν1/ν, which the marginal leaves unidentified; this yields a policy-relative observability variance that measures dependence on visible evidence, distinct from variation induced by resampling the training set. On a Gaussian-process benchmark with a contextual reference built from analytic Bayes conditional means, MORSE has the lowest log-variance error among compared evidential, ensemble, and dropout estimators in all nine tested settings, while the estimator closest to training-set variation perturbs the model rather than its context in every setting, so the two references separate the estimators empirically. MORSE amortizes observability estimation into one forward pass per observation state. On a heteroscedastic pulse benchmark, its predictor's cross-context variance approaches a large-sample reference as training data increase. MORSE also improves low-label transfer on PTB-XL and in-distribution trajectory prediction on IMPTC.
Oct 29, 2026

Integrated Feature Selection for classification and Representation Learning with Group-Lasso Siamese Networks

The Least Absolute Shrinkage and Selection Operator (Lasso) has been widely employed for feature selection in linear models and has also been extended to multi-layer perceptron (MLP) networks for both feature and group feature selection. Group-Lasso has also been used in feature selection using feed-forward neural networks. To the best of our knowledge, this is the first study to incorporate the Group-Lasso principle into a Siamese network architecture for feature selection. Unlike existing approaches that primarily optimize classification accuracy, the proposed method aims to identify a compact subset of features that simultaneously preserves the intrinsic representation of the input data (the goal of Siamese network) and is expected to achieve high classification performance. Two variants of the proposed framework are developed: one based on the Euclidean distance and the other on a modified Jaccard similarity measure for real-valued vectors. The latter is specifically designed to improve feature selection for high?dimensional sparse data. The effectiveness of the proposed approach is demonstrated on 12 benchmark datasets with feature dimensions ranging from 6 to 7070 using eight different classifiers. Experimental results demonstrate that the proposed Siamese Lasso framework consistently identifies compact and discriminative feature subsets maintaining competitive or superior classification performance, although the loss function used explicitly did not consider any regularizer to learn classification. In addition, the quality of the embeddings produced by these networks is assessed using manifold learning techniques such as UMAP and t-SNE.
Oct 29, 2026

PRISM: Personalized Item Representations in User History

Sequential recommenders typically construct history inputs from item identifiers or content features using an embedding map shared across users. The same item therefore yields the same initial embedding representation, even when users have different relationships to the item. We propose personalized item representations in user history, where each item's embedding expresses its relationship to the user. Personalization becomes a property of the history inputs themselves, supplying richer evidence for later sequence modeling. We instantiate this idea with PRISM, which reuses upstream ranking computation to generate personalized item embeddings. An autoencoder jointly compresses all layer states from a user tower, a user--item tower, and same-request candidate mixing into one embedding per scored pair for downstream replay. Offline evaluation on Surface 1, a major recommendation surface at billion scale, reports an average downstream normalized entropy (NE) change of −0.15%-0.15\%, compared with −0.07%-0.07\% for VISTA at similar history lengths. We further identify an empirical sequence-length scaling law: over history lengths of 26--240 events, NE gain follows a fitted power law with exponent 0.468, increasing from 0.06% to 0.17%. Therefore, personalized item representations in user history provide a new scaling dimension for improving recommendation quality.
Oct 28, 2026

FairSplit: Decomposing the Embedding Space for Fair Classification

Fair classification faces a twofold challenge: maintaining high classification accuracy while limiting reliance on sensitive information. Our method FairSplit achieves this by learning a decomposition of the embedding space into a Fair Space used for prediction and a Sensitive Space that isolates sensitive information. During training, the model must simultaneously learn the classification task and separate fair from sensitive information. The Sensitive Space gives sensitive attributes a destination, so the Fair Space can be used for prediction without relying on them. This decomposition is guided via a learnable mask and the Hilbert-Schmidt Independence Criterion (HSIC) as a measure of statistical dependence. We provide theoretical guarantees showing that the demographic parity gap is upper-bounded by both (i) the learned dimensionality of the Fair Space and (ii) the HSIC between the Fair Space and sensitive attributes. FairSplit effectively handles multiclass tasks and non-binary sensitive attributes, overcoming key limitations of existing fair classification methods. Experiments on established fairness benchmarks show that FairSplit achieves competitive accuracy-fairness trade-offs. Code is available at https://github.com/shkabrii/fairsplit.
Oct 28, 2026

SpectralTF-MAE: A Pre-trained Spectral Foundation Model for Tokamak Fusion Plasma Diagnostics

Spectroscopic diagnostics provide rich information for characterizing plasma states in tokamak fusion experiments, while existing learning-based approaches are often designed for individual tasks and rely heavily on labeled data. In this work, we propose SpectralTF-MAE, a self-supervised spectral foundation model for tokamak plasma diagnostics. The model is pretrained on paired charge-exchange recombination spectroscopy (CXRS) and X-ray crystal spectroscopy (XCS) measurements through masked spectral reconstruction. To better capture diagnostically important spectral structures, we introduce a peak-aware masking strategy together with a peak-enhanced reconstruction objective, encouraging the model to focus on characteristic spectral regions. A shared Transformer encoder is employed to learn transferable representations across different diagnostic modalities. The pretrained model is further adapted to several downstream tasks, including plasma-parameter prediction, cross-diagnostic spectral generation, and spectral peak analysis. Experiments on EAST diagnostic data demonstrate that the proposed framework achieves competitive performance across different tasks and consistently benefits from self-supervised spectral pretraining. These results indicate the potential of SpectralTF-MAE as a general and transferable representation learning framework for plasma spectroscopic diagnostics. The source code and pre-trained models will be released on https://github.com/Event-AHU/OpenFusion/tree/main/SpectralTF-MAE.
Oct 28, 2026

Leaving the Cave by Breaking the Symmetry: Identifiable Structure Can Enable Cross-Lingual Generalization

Why do multilingual language models generalize across languages even without shared vocabulary? The Platonic Representation Hypothesis posits that models trained on different data, objectives, and modalities converge toward a shared statistical model of reality. Extending this intuition to representations inside a single multilingual model, we propose the Relational Identifiability Hypothesis: relations in a shared world can promote entity alignment by ruling out alternative cross-lingual correspondences, making corresponding entities more likely to become directly readable without an additional transformation. We test this hypothesis in a controlled bilingual Transformer whose two token-disjoint languages describe the same modular-arithmetic world. We find that the languages can develop a shared answer subspace without aligning corresponding numbers. Cross-lingual alignment emerges selectively when the available relations constrain number correspondences: subtraction makes zero task-level identifiable, inducing selective alignment, while an order-sensitive operation produces broad alignment across numbers. These patterns predict success in zero-shot cross-lingual transfer. Together, our results show that relational structure in a shared world can promote entity alignment and enable transfer.
Oct 27, 2026

Investigating the Early Signatures of Feature Learning in Deep Linear Networks

Deep neural networks learn data-dependent features, a capacity widely credited for their empirical success. However, explicitly characterizing what features emerge -- and when they improve generalization -- is challenging, since training is nonlinear and non-convex and the data is often anisotropic. Here, we give a closed-form account of feature learning for a deep linear network in a mean-field parameterization, trained by one gradient step from initialization on high-dimensional data carrying a low-dimensional (spiked) signal. In the infinite-width mean-field limit, one step updates each layer's feature kernel by a target-aligned rank-one spike, amplified by both the learning rate and the layer depth. We next assess how these features impact prediction through kernel ridge regression (KRR), finding that KRR on the learned feature kernel yields scalar inflation of the KRR estimator at initialization -- an "anti-shrinkage" mechanism that reinflates task-relevant directions suppressed by ridge, while also amplifying noise. To study when this inflation improves generalization, in the overparameterized proportional limit we derive closed-form expressions for the bias, variance, and generalization errors of these estimators in terms of key macroscopic parameters -- notably the spike strength and spike-target alignment. Crucially, we find that a one-step generalization advantage vanishes for isotropic inputs and appears when both the spike strength and an effective signal-to-noise ratio are sufficiently large, demonstrating that a successful first gradient step both requires and benefits from low-dimensional task structure.
Oct 27, 2026

Dynamic Merging Mutually-Distilled Multi-View Networks for RGB-Event-based Representation Learning

Multi-modal fusion algorithms combining conventional frame-based cameras and event cameras have recently attracted substantial research attention. Typically, existing approaches convert event streams into frames, voxels, or tensors, and then learn the required spatiotemporal information from these transformed event representations for fusion with conventional frame-camera data. Obviously, such paradigms only exploit the limited representational capacity of event data. Moreover, standard training pipelines fail to fully leverage diverse model parameters obtained throughout the learning process. In this paper, we propose a novel mutual distillation method leveraging multi-view teacher network weights to better integrate multiple event representations and achieve model-parameter level fusion. Specifically, we employ multiple encoders to extract different types of event representations and perform mutual knowledge distillation at both the feature-level and logit-level in this process. Afterwards, we perform a dynamic hierarchical model merge strategy on the model parameters from the first stage to yield a unified multi-modal RGB-Event representation model. Experimental results across multiple event-based datasets demonstrate that our proposed framework achieves significant performance improvements.
Oct 26, 2026

What Should an Information Bottleneck Encoder Learn? Separating Representation Learning from Rate Control

The information bottleneck (IB) trades compression of an input XX against prediction of a target YY through a multiplier β=1/λβ=1/λ, yet each operating point is usually served by training, or fully conditioning, a new encoder. For jointly Gaussian data, we show that the IB solution factorizes as A(β)=D(β)WA(β)=D(β)W: the information-bearing directions WW do not depend on ββ, and ββ acts only through scalar gains. We ask whether this separation of representation from rate control survives beyond Gaussian models, treating the budget as a given operating parameter rather than a quantity to be optimized. For a budget-free representation F(X)F(X), the separation gap is controlled by the sufficiency deficit εF=I(X;Y∣F(X))ε_F=I(X;Y\mid F(X)): We show there is at least εFε_F in the worst case for finite inputs, and it is bounded uniformly in λλ by a function of εFε_F for finite label alphabets. The same certificate extends to a new target distribution and to pooled multi-domain data, where a conditional-alignment term appears. On MNIST, SVHN, CIFAR-10 and STL-10 (10 seeds, eight prespecified budgets), a frozen CNN representation with one compressor--decoder pair per budget reduces estimated training multiply--accumulate operations (MACs) by 7.7×7.7\times and wall-clock time by about 2×2\times relative to end-to-end VIBs, and attains higher budget-averaged accuracy (65.6%65.6\% versus 57.3%57.3\% per domain; 67.1%67.1\% versus 57.1%57.1\% pooled), with the advantage concentrated at strong compression, where the VIBs collapse. At weak compression (λ≤0.036λ\le0.036), VIB has higher cross-entropy and ECE; its surrogate Lagrangian is higher per domain, indistinguishable from the pooled VIB in the macro average, and lower on the pooled test set.
Oct 26, 2026

A theory of noisy linear recurrent neural network learning dynamics

Biological neural networks are noisy: their dynamics involve difficult-to-predict fluctuations which are likely due not just to chaos, but also to a variety of low-level biophysical processes. Despite the omnipresence of intrinsic noise, its theoretical consequences are incompletely understood, especially in the context of learning. For example, does noise meaningfully shape representation learning, or is it better viewed as a nuisance factor whose presence or absence changes little about what solution is found and when? Here, we argue that intrinsic noise can significantly modulate representation learning in recurrent neural networks (RNNs), since it can strongly break solution-space degeneracies and in some cases control whether a non-degenerate solution exists at all. We show that this is true for an analytically tractable model of noisy RNN learning with a linear activation function and white noise inputs. We find that, in the so-called rich regime (where weights are initialized with small values), noise strongly modulates the directions in which the read-in, readout, and recurrent weights align. Moreover, we find that noise produces a kind of phase transition: for sufficiently high noise, a non-degenerate solution is not even optimal if parameter scales are limited (e.g., via regularization). These results extend recent analytic work on RNN representation learning and show that noise can act as a fundamental determinant of representations.
Oct 26, 2026

AutoPRL: Automated Path Representation Learning via a Learnable View Generator

Self-supervised learning has emerged as a powerful paradigm for Path Representation Learning (PRL) in road networks. PRL aims to learn informative representations of paths traversed by moving objects, such as vehicles, where each path consists of a sequence of connected road segments. These representations are essential for various downstream tasks, including travel-time estimation and path ranking. However, existing methods often suffer from ineffective view generation and inefficient training procedures due to their reliance on heuristic augmentations and computationally intensive contrastive learning. To address these issues, we propose AutoPRL, an effective and efficient self-supervised PRL framework that integrates three key components. The Automatic Data Augmentation (AutoDA) module employs variational autoencoders to generate semantically meaningful and diverse feature-level path views by modeling each topology-aware node embedding with a Gaussian latent distribution, thereby mitigating information loss and enhancing representation quality. Second, the Cross-View Decoupled Contrastive (CVDC) learning mechanism improves training efficiency by disentangling the optimization of positive and negative pairs across node- and path-level views, enabling more stable contrastive learning. Third, an Adaptive Co-Training (ACT) module dynamically adjusts uncertainty-based loss weights to balance multiple objectives, ensuring efficient and scalable training. Extensive experiments on three real-world datasets demonstrate that AutoPRL achieves strong and consistent performance in both effectiveness and efficiency across multiple downstream tasks.
Oct 24, 2026

Sliding Networks: A Family of Residual Structures Preserving Feature Learning and Feature Diversity

Deep learning models tend to perform best at large widths and depths. The growing interest in scaling laws, especially in the language modeling domain, has motivated procedures for scaling up network widths and depths in order to preserve stability and computational efficiency of training dynamics. Existing depth-scaling approaches force a choice between retaining locally maximal feature learning and preserving a form of feature diversity in the residual stream arising from random differences between channels. We examine a joint scaling recipe that preserves both. With residual width N♭N_\flat, hidden width N♯N_\sharp, and depth LL, the key quantity is, by default, the limiting ratio N♭/(N♯L)→αN_\flat/(N_\sharp L)\toα, while correlated initialization (Half Copy) can partially relax the resulting shape constraints. Theory and residual-MLP experiments support convergence at fixed positive αα to an SDE limit with locally maximal feature learning. Hidden computations can then be redistributed between width and depth while preserving this limit, like turning melody into harmony and back. Language-model experiments nevertheless show performance differences under regrouping at fixed parameter count. These results distinguish approximating a limit (ODE or SDE) from scaling for better performance.
Oct 24, 2026

Greedy Orthogonal Polynomial Manifolds with Lipschitz Control for Dimensionality Reduction and Reduced Modeling

Polynomial manifolds provide nonlinear low-dimensional representations through polynomial decoders. We propose to represent the polynomial terms of the decoder using features that are orthonormal on the training data. This empirical orthogonality enables a greedy algorithm, Greedy-OPM, which jointly selects encoder modes and polynomial features while minimizing the reconstruction error under a prescribed bound on the decoder's Lipschitz constant. Numerical experiments show that this Lipschitz constraint improves generalization away from the training data while retaining high reconstruction accuracy.
Oct 24, 2026

Rethinking Microscopy Generation: Co-Designed Diffusion for Biologically Interpretable Single-Cell Synthesis

Generative models for microscopy can already produce realistic whole-image fields of view, but offer no principled way to control how each individual cell looks. We address this gap by recasting microscopy generation at the single-cell level and designing the input controls and the model's internal representation jointly, so that both are organized along the same biologically meaningful directions. On the input side, we aggregate a long list of standard cell measurements into a small set of semantic axes (size, elongation, brightness, texture, etc.) that are each named, measurable, and traversable. On the model side, we introduce BioVAE, an encoder-decoder whose latent space is explicitly aligned with the same set of measurements. A key finding is that the benefit of biologically structuring the latent emerges only when conditioning is also moved to the semantic axes, showing that the two ends must be designed jointly. Experiments on LIVECell show that semantic-axis conditioning substantially improves generative fidelity over high-dimensional conditioning, a blind study with biologists confirms the realism of generated cells, and per-axis traversals produce stable and interpretable phenotypic changes.
Oct 23, 2026

Exploring Joint-Embedding Predictive Architectures for Self-Supervised Representation Learning of Powder X-ray Diffraction Patterns

Powder X-ray diffraction (XRD) is the primary high-throughput method for resolving crystal structures, a cornerstone of materials science. As high-throughput synthesis and self-driving laboratories advance, automated XRD pattern analysis has become vital for accelerating materials discovery. However, traditional matrix factorization methods fail when lattice variations cause non-linear peak shifts. At the same time, supervised learning approaches face a combinatorial explosion when handling multi-phase mixtures. To overcome these challenges, we adopt a self-supervised Joint Embedding Predictive Architecture (JEPA) to learn representation embeddings from 1D diffraction patterns. By optimizing a masked self-prediction loss directly within the latent space, the framework captures high-level crystallographic semantics. Downstream evaluations show that lightweight attentive probes trained on frozen JEPA representations achieve space group classification and lattice parameter estimation accuracy close to much larger supervised vision transformers. Furthermore, latent probing reveals that the learned representation space naturally organizes physical transformations, isolating smooth trajectories for peak broadening. Interestingly, decoding along latent paths between single-phase crystal representations yields realistic multi-phase mixture patterns, demonstrating an emergent continuous structure learned without explicit multi-phase supervision.
Oct 23, 2026

Closing the Modality Gap via Simplex-Constrained Representations

While multimodal contrastive models like CLIP and CLAP successfully learn shared representations, their embeddings typically exhibit a persistent modality gap. We propose addressing this separation through a purely geometric approach: changing the embedding domain from the unit sphere, Sd−1S^{d-1}, to the probability simplex, Δd−1Δ^{d-1}. Although both manifolds share the same intrinsic dimension, the simplex acts as a structural prior; its nonnegative, unit-mass constraints serve as a cross-modal calibration that forces both modalities to ground themselves in the same coordinate system. By attaching lightweight softmax adapters and using Total Variation similarity, we reduce the centroid ℓ2\ell_2 gap by 97--99% across five multimodal retrieval benchmarks. This confirms that simplicial geometry induces genuine semantic alignment, matching or exceeding the retrieval performance of standard unconstrained models.
Oct 23, 2026

PlayerEmbed: Learning Behavioral Fingerprint Representations from Raw Input Sequences

Raw player input sequences in video games contain fine-grained behavior, but are high-dimensional and noisy, which makes large-scale analytics challenging. To address this, we introduce PlayerEmbed, a framework that learns compact and discriminative player embeddings using deep metric learning. Our approach uses a hybrid CNN-LSTM encoder trained on raw input sequences from 1,000 players. We evaluate the model on held-out sequences from these players and on sequences from 1,000 players entirely unseen during training to assess generalization to unseen identities. For player identification, a nearest-centroid classifier built from reference embeddings achieves strong performance, reaching balanced accuracies of 0.950 for players seen during training and 0.879 for players unseen during training. Systematically increasing the number of training players substantially improves generalization to unseen players, indicating a transition from simple fingerprinting to learning transferable behavioral patterns.
Oct 23, 2026

Shared Is Not Enough: Predictor Width and Factor Readability Under Masked Latent Prediction

Masked latent prediction methods such as I-JEPA aim to learn predictable features, but what determines which information they retain is less clear. We built a simulator in which every factor behind an image is known, so a probe can measure how readable each factor remains after training. We trained small encoders (0.7M parameters) with I-JEPA, SimCLR and BYOL on the same pairs of views and probed a factor that both views share. I-JEPA had a predictor with half the encoder's width and depth, the proportions it uses for ViT-B. Its median score was about 40% of SimCLR's. Widening the predictor to the encoder's width reversed this ordering at every training seed, against both other methods. We release the simulator and the trained models.
Oct 23, 2026

Beyond Channel Data: Configuration-Enriched Radio Foundation Models

This paper proposes incorporating channel configuration (CFG) information into Radio Foundation Model (RFM) pre-training and demonstrates that the resulting CFG-enriched embeddings serve as effective inductive priors for downstream tasks under label scarcity. We compare the proposed embeddings with raw channel tokens and multiple baseline models, under three different data regimes, for multiple downstream tasks --- Line-of-Sight/Non-Line-of-Sight (LoS/NLoS) classification, timing advance (TA) estimation, and beam selection. Evaluation is done in both in-distribution and out-of-distribution configuration settings. Experimental outcomes show that the proposed embeddings consistently outperform and/or are competitive to the baselines, across experiments in both seen and unseen configuration scenarios.
Oct 22, 2026

Rethinking Multi-View Fusion for Urban Representation Learning

Urban region representation learning encodes heterogeneous urban signals, such as mobility flows, points of interest, and land-use patterns, into unified embeddings for downstream tasks, including check-in prediction, crime forecasting, and service demand estimation. Existing approaches rely on globally shared fusion and joint optimization, implicitly assuming uniform modality importance and naturally balanced multi-objective training. However, urban data is inherently heterogeneous. Our empirical analysis reveals two structural issues: (i) modality utility varies markedly across regions, causing static fusion to overemphasize signals from certain modalities; and (ii) heterogeneous objectives exhibit persistent gradient imbalance, causing representational bias and unstable training. We present PRISM, a self-adjusting framework for adaptive multi-view learning. PRISM adapts sparse Mixture-of-Experts (MoE) routing to learn region-specific view allocations, while a Proportional-Integral-Derivative (PID)-inspired regulation mechanism mitigates gradient conflicts during optimization. Experiments on three large-scale urban datasets show that PRISM achieves more competitive performance than strong multi-view baselines, while providing more stable optimization behavior and more balanced region representations.
Oct 22, 2026

Early learning in large residual networks

We study the early training dynamics of nonlinear residual networks with blocks of zero hidden layers as a model of more general residual networks. We analyze networks in a sequential limit where first width and depth diverge, then the number of training data points diverges, and finally the input dimension diverges. This is a more refined setup compared to the typical mean-field theory analysis in which depth and width diverge but the number of training datapoints and input dimension are fixed. Our main technical result is a diagrammatic computation of both the network output and the hidden representations after the few steps of gradient descent in terms of the statistics of inner products of inputs from the training set. One of our principle conceptual findings is a negative result: in the limit of large input dimension the network output can learn only a linear function of its inputs early in training since higher learned monomials are more and more suppressed in the input dimension and vanish in the large dimension limit. This suggests that mean-field limits of ResNets are surprisingly rigid in the regime of high input dimension and may not be sufficient toy models for understanding feature learning. Finally, we investigate how to ensure hyperparameter transfer for the learning rate as a function of dataset size and input dimension. We test our predictions numerically and find excellent agreement with our theory.
Oct 22, 2026

GigaBench: Benchmarking Learnable Representations for Gigapixel Image Compression and Super-Resolution

Learnable Representations (LRs) can fit parametric families of functions to diverse kinds of signals. These emergent neural-native data formats are gathering interest in the scientific community due to their compactness, resolution-agnosticism and amenability to direct neural processing. On the other hand, gigapixel images are becoming increasingly widespread across a variety of scientific, cultural, and industrial domains. In this paper, we present GigaBench, the first benchmark and experimental evaluation focused on deploying LRs with gigapixel imagery. GigaBench includes images collected, acquired, and synthetically rendered across diverse domains, and addresses image compression and super-resolution. We adapt a representative set of LRs, spanning Implicit, Hybrid, and Explicit methods, so as to operate reliably at the gigapixel scale on consumer hardware. Through systematic evaluation across reconstruction quality, model size, compression and inference times, we identify consistent trends across methods and suggest actionable guidance for future investigation.
Oct 22, 2026

EEG-Captioner: Grounding EEG-to-Language Generation in Fundamental Signal Properties

Electroencephalography (EEG)-to-language models offer a promising approach for translating complex neural signals into human-readable descriptions. However, generating plausible language does not necessarily demonstrate that a model has learned to interpret the underlying EEG signal. A direct way to assess such interpretation is to probe objectively measurable properties of the signal itself. We therefore construct an EEG-language dataset of 160,781 segments annotated with signal-level question answering (QA) pairs, where each question targets one of five property families and requires a quantitative comparison involving spectral, spatial, temporal, or lateral signal properties. Each segment is also paired with a feature-grounded EEG caption. We further propose a Dual-Domain EEG Encoder that jointly models temporal waveforms and spectral-spatial representations, and introduce EEG-Captioner, a unified EEG-to-language framework trained with both signal-level QA and EEG caption supervision. Experiments show that the proposed encoder achieves the highest overall macro-averaged balanced accuracy (Macro-BA) among the compared EEG encoders, while EEG-Captioner achieves a Macro-BA of 88.81% on signal-level QA and improves caption factual consistency from 82.06% to 83.46% over caption-only language adaptation, demonstrating the complementary benefits of signal-level QA and caption supervision. Demo page: https://lyJinAoki.github.io/EEG-Captioner/.
Oct 21, 2026

Optimal Multitask Representations Under a Rate Constraint: An Information Bottleneck Analysis

Which pretraining objective yields the best representation for downstream tasks under an information budget? We formalize this problem as a multitask information bottleneck, using a mutual information constraint to model incomplete information retention. For Gaussian inputs and tasks, we characterize optimal representations and show that a weakly task-aligned specialist can outperform a reconstruction generalist at low rates but lose at high rates. We also construct task families for which every fixed Gaussian prediction objective incurs arbitrarily large relative regret, driven by reversals in optimal allocation priorities. Conversely, the average task information matrix defines an objective with a uniform relative-regret bound controlled by maximum task SNR. CelebA and dSprites experiments illustrate that compression can change which pretraining objective performs best, and even task-aware training can reverse which downstream task is better preserved as rate increases.