Searcharxiv⌕ Search

arXiv subjects

Gal Mishne

Publications and source records attributed to Gal Mishne.

At least 19 recordsLinked to original sources

Representational Geometry as a Diagnostic for Adversarial Robustness

Adversarial training is a common technique for making neural networks more robust to small, intentionally crafted input perturbations, and robustness is commonly measured through accuracy under attack. By exploring inter-network representations across different adversarial training objectives, we can analyze whether multiple networks organize their latent representations similarly. We study Multilayer Perceptrons trained on MNIST under three adversarial training regimes and compare the resulting population training dynamics and geometry to those of a vanilla network. We use M-PHATE to embed data samples across training epochs and examine how the geometry of the data changes during training under adversarial objectives. On penultimate-layer representations we measure displacement between clean and perturbed points, distributional distance through maximum mean discrepancy, and nearest-neighbor recovery. Our visualizations and quantitative analyses show that adversarial training leads to structural changes in how networks organize internal data representations, and robust accuracy and representational stability can rank the same defenses in opposite ways. These findings support a discussion of re-evaluating how practitioners approach the auditing of model robustness by treating latent representations and their training trajectories as diagnostic artifacts across sets of networks.

cs.LG↗

Beyond Flat Walks: Compositional Abstraction for Autoregressive Graph Generation

Autoregressive graph generators typically linearize graphs through flat node-edge walks, which can obscure higher-order structure in large and complex graphs. We introduce a hierarchical graph abstraction and coarse-to-fine generation pipeline that separates coarsening, tokenization, and decoding. This exposes three design axes: community coarsening, hierarchical tokenization, and order randomization. Across these axes, traversal-based hierarchy consistently improves distributional and structural fidelity, while its V.U.N.\ cost depends strongly on the coarsening and randomization policy. In particular, the fidelity--V.U.N.\ trade-off observed under within-community randomization is substantially reduced when traversal order is randomized across the full hierarchy. With a fixed backbone, training protocol, and matched training time across five molecular and five non-molecular benchmarks, hierarchical serialization yields consistent distributional and structural gains on all molecular datasets, while the non-molecular results further show that these gains depend on how well the supplied hierarchy aligns with structure in the target distribution. Together, these results characterize when and how hierarchical serialization benefits autoregressive graph generation.

cs.LG↗

NearFormer: Linear-Size Sparse Attention via Randomized Near Neighbors

Graph transformers can capture long-range interactions, but their quadratic attention cost limits their use on large graphs. We introduce \emph{NearFormer}, a graph transformer whose global attention is restricted to a randomized near-neighbor (NN) graph. NearFormer uses an efficient randomized construction to connect each node to a small number of near neighbors from a broader neighborhood defined by a label-independent feature or spectral geometry. For a fixed retained budget \(k\), the resulting attention support has linear size \(\mathcal{O}(|E|+nk)\). The NN graph is built once by within-bin search and then held fixed. We use these edges as sparse global-attention supports, with optional learned attention biases, while preserving the original graph for local message passing. Experiments on long-range graph benchmarks and synthetic structural tasks show that NearFormer provides a favorable accuracy--scalability trade-off relative to local message passing, dense and sparse attention.

cs.LG↗

Robust Tangent Space Estimation via Laplacian Eigenvector Gradient Orthogonalization

Estimating the tangent spaces of a data manifold is a fundamental problem in geometric data analysis. The standard approach, Local Principal Component Analysis (LPCA), struggles in high-noise settings due to a critical trade-off in choosing the neighborhood size. Selecting an optimal size requires prior knowledge of the geometric and noise characteristics of the data that are often unavailable. In this paper, we propose a spectral method, Laplacian Eigenvector Gradient Orthogonalization (LEGO), that utilizes the global structure of the data to guide local tangent space estimation. Instead of relying solely on local neighborhoods, LEGO estimates the tangent space at each data point by orthogonalizing the gradients of low-frequency eigenvectors of the graph Laplacian. We provide two theoretical justifications of our method. First, a differential geometric analysis on the tubular neighborhood of a manifold shows that gradients of the low-frequency Neumann eigenfunctions of the tube align closely with the manifold's tangent bundle, while an eigenfunction with high gradient in directions orthogonal to the manifold lies deeper in the spectrum. Second, a random matrix theoretic analysis also demonstrates that low-frequency eigenvectors are robust to sub-Gaussian noise. These results allow us to derive the asymptotic scaling and stability of the estimated eigenvector gradients. Numerical experiments demonstrate that LEGO yields tangent space estimates that are significantly more robust to noise than those from LPCA, resulting in marked improvements in downstream tasks such as manifold learning, boundary detection, and local intrinsic dimension estimation.

cs.LG↗

NEST: Nested Embedding Representation for Multiscale Clustering

Real-world data often exhibits cluster structure at multiple resolutions, yet existing methods typically learn a single partition or construct an explicit hierarchy. Inspired by Matryoshka Representation Learning, we introduce NEST, an unsupervised framework that encodes coarse-to-fine structure within a single embedding. NEST jointly optimizes clustering and separation across nested prefix subspaces, enabling progressively finer distinctions as embedding dimensionality increases, with rank and within-cluster regularization that keep the embedding from collapsing. Empirically, distances in the learned embedding follow the hierarchy distance more closely than those of the input, although the hierarchy is never observed during training. On synthetic hierarchies, NEST recovers multiscale structure given only the number of clusters per level. Experiments on vision, text, and single-cell data further demonstrate multiscale organization across real-world representations. By making clustering granularity an intrinsic property of the representation, NEST provides a unified approach to multiscale structure discovery.

cs.LG↗

Neural Algorithmic Reasoning for Graph Saddle Point Problems

Neural algorithmic reasoning, or aligning a neural network with an algorithmic paradigm, has emerged as an approach to solving polynomial-time-solvable and computationally harder combinatorial optimization problems. We propose a new message-passing framework based on the Chambolle-Pock Primal--Dual Hybrid Gradient (PDHG) method called \textsc{GraphPDHG} for solving general graph saddle-point problems. Theoretically, we show that \textsc{GraphPDHG} can efficiently solve a family of graph saddle-point problems by simulating PDHG. We also show that our network can learn an accelerated PDHG algorithm. Experimentally, we support our results on accelerated PDHG by evaluating the performance of our model as a learned warm start for second-order optimization techniques (SSNAL). We also show that alignment with PDHG leads to stronger size generalization than non-aligned graph neural network (GNN) baselines. Overall, we propose a novel architecture for solving a general family of optimization problems on graphs.

cs.LG↗

Reassessing Muon for Matrix Factorization

Muon has recently emerged as a strong optimizer for large-scale deep learning, where it reshapes gradient updates through approximate orthogonalization and has been reported to outperform Adam and AdamW in large language model training. Its empirical success has motivated a growing body of theoretical work that interprets Muon as steepest descent under the spectral norm. Yet it remains unclear which of Muon's advantages stem from its update rule itself and which are artifacts of the scale, architecture, and data of modern deep networks. In this work, we isolate the optimizer from these confounding factors by studying Muon on a simple, well-understood, and spectrally structured problem: low-rank matrix factorization. Through a controlled comparison against carefully tuned adaptive baselines, we find that Muon does not consistently outperform AdamW in this setting and that several previously reported advantages are sensitive to hyperparameter choices. Our results provide a more nuanced picture of when spectrum-aware orthogonalization is beneficial and argue for evaluating modern optimizers on controlled problems in addition to end-to-end benchmarks.

cs.LG↗

Cluster and then Embed: A Modular Approach for Visualization

Dimensionality reduction methods such as t-SNE and UMAP are popular methods for visualizing data with a potential (latent) clustered structure. They are known to group data points at the same time as they embed them, resulting in visualizations with well-separated clusters that preserve local information well. However, t-SNE and UMAP also tend to distort the global geometry of the underlying data. We propose a more transparent modular approach that first clusters the data, then embeds each cluster, and finally aligns the clusters to obtain a global embedding. We demonstrate this approach on several synthetic and real-world datasets and show that it is competitive with existing methods, while being much more transparent.

cs.LG↗

Robust boundary detection and density estimation using doubly stochastic scaling of the Gaussian kernel

This paper addresses the problem of detecting boundary points and estimating the sampling density of a dataset derived from a compact manifold with boundary, potentially in the presence of noise. We extend recent advances in doubly stochastic scaling of the Gaussian heat kernel via Sinkhorn iterations to this setting. Our main contributions are: (a) deriving a characterization of the scaling factors for manifolds with boundary, (b) developing a boundary direction estimator aimed at identifying boundary points followed by a boundary-corrected kernel density estimator based on doubly stochastic kernel and local principal component analysis, and (c) demonstrating through simulations that the resulting estimates of the boundary points and the sampling density outperform the standard Gaussian kernel-based approach, particularly under noisy conditions.

math.ST↗

Multiway Multislice PHATE: Visualizing Hidden Dynamics of RNNs through Training

Recurrent neural networks (RNNs) are a widely used tool for sequential data analysis; however, they are still often seen as black boxes. Visualizing the internal dynamics of RNNs is a critical step toward understanding their functional principles and developing better architectures and optimization strategies. Prior studies typically emphasize network representations only after training, overlooking how those representations evolve during learning. Here, we present Multiway Multislice PHATE (MM-PHATE), a graph-based embedding method for visualizing the evolution of RNN hidden states across the multiple dimensions spanned by RNNs: time, training epoch, and units. Across controlled synthetic benchmarks and real RNN applications, MM-PHATE preserves hidden-representation community structure among units and reveals training-phase changes in representation geometry. In controlled synthetic systems spanning multiple bifurcation families and smooth state-space warps, MM-PHATE recovers qualitative dynamical progression while distinguishing family-level differences. In task-trained RNNs, the embedding identifies information-processing and compression-related phases during training, and time-resolved geometric and entropy-based summaries align with linear probes, time-step ablations, and label--state mutual information. These results show that MM-PHATE provides an intuitive and comprehensive way to inspect RNN hidden dynamics across training and to better understand how model architecture and learning dynamics relate to performance.

cs.LG↗

ALIGN: Adversarial Learning for Generalizable Speech Neuroprosthesis

Intracortical brain-computer interfaces (BCIs) can decode speech from neural activity with high accuracy when trained on data pooled across recording sessions. In realistic deployment, however, models must generalize to new sessions without labeled data, and performance often degrades due to cross-session nonstationarities (e.g., electrode shifts, neural turnover, and changes in user strategy). In this paper, we propose ALIGN, a session-invariant learning framework based on multi-domain adversarial neural networks for semi-supervised cross-session adaptation. ALIGN trains a feature encoder jointly with a phoneme classifier and a domain classifier operating on the latent representation. Through adversarial optimization, the encoder is encouraged to preserve task-relevant information while suppressing session-specific cues. We evaluate ALIGN on intracortical speech decoding and find that it generalizes consistently better to previously unseen sessions, improving both phoneme error rate and word error rate relative to baselines. These results indicate that adversarial domain alignment is an effective approach for mitigating session-level distribution shift and enabling robust longitudinal BCI decoding.

cs.LG↗

MINAR: Mechanistic Interpretability for Neural Algorithmic Reasoning

The recent field of neural algorithmic reasoning (NAR) studies the ability of graph neural networks (GNNs) to emulate classical algorithms like Bellman-Ford, a phenomenon known as algorithmic alignment. At the same time, recent advances in large language models (LLMs) have spawned the study of mechanistic interpretability, which aims to identify granular model components like circuits that perform specific computations. In this work, we introduce Mechanistic Interpretability for Neural Algorithmic Reasoning (MINAR), an efficient circuit discovery toolbox that adapts attribution patching methods from mechanistic interpretability to the GNN setting. We show through two case studies that MINAR recovers faithful neuron-level circuits from GNNs trained on algorithmic tasks. Our study sheds new light on the process of circuit formation and pruning during training, as well as giving new insight into how GNNs trained to perform multiple tasks in parallel reuse circuit components for related tasks. Our code is available at https://github.com/pnnl/MINAR.

cs.LG↗

Multi-Integration of Labels across Categories for Component Identification (MILCCI)

Many fields collect large-scale temporal data through repeated measurements (trials), where each trial is labeled with a set of metadata variables spanning several categories. For example, a trial in a neuroscience study may be linked to a value from category (a): task difficulty, and category (b): animal choice. A critical challenge in time-series analysis is to understand how these labels are encoded within the multi-trial observations, and disentangle the distinct effect of each label entry across categories. Here, we present MILCCI, a novel data-driven method that i) identifies the interpretable components underlying the data, ii) captures cross-trial variability, and iii) integrates label information to understand each category's representation within the data. MILCCI extends a sparse per-trial decomposition that leverages label similarities within each category to enable subtle, label-driven cross-trial adjustments in component compositions and to distinguish the contribution of each category. MILCCI also learns each component's corresponding temporal trace, which evolves over time within each trial and varies flexibly across trials. We demonstrate MILCCI's performance through both synthetic and real-world examples, including voting patterns, online page view trends, and neuronal recordings.

cs.LG↗

Unsupervised Feature Selection Through Group Discovery

Unsupervised feature selection (FS) is essential for high-dimensional learning tasks where labels are not available. It helps reduce noise, improve generalization, and enhance interpretability. However, most existing unsupervised FS methods evaluate features in isolation, even though informative signals often emerge from groups of related features. For example, adjacent pixels, functionally connected brain regions, or correlated financial indicators tend to act together, making independent evaluation suboptimal. Although some methods attempt to capture group structure, they typically rely on predefined partitions or label supervision, limiting their applicability. We propose GroupFS, an end-to-end, fully differentiable framework that jointly discovers latent feature groups and selects the most informative groups among them, without relying on fixed a priori groups or label supervision. GroupFS enforces Laplacian smoothness on both feature and sample graphs and applies a group sparsity regularizer to learn a compact, structured representation. Across nine benchmarks spanning images, tabular data, and biological datasets, GroupFS consistently outperforms state-of-the-art unsupervised FS in clustering and selects groups of features that align with meaningful patterns.

cs.LG↗

Coupled Transformer Autoencoder for Disentangling Multi-Region Neural Latent Dynamics

Simultaneous recordings from thousands of neurons across multiple brain areas reveal rich mixtures of activity that are shared between regions and dynamics that are unique to each region. Existing alignment or multi-view methods neglect temporal structure, whereas dynamical latent variable models capture temporal dependencies but are usually restricted to a single area, assume linear read-outs, or conflate shared and private signals. We introduce the Coupled Transformer Autoencoder (CTAE) - a sequence model that addresses both (i) non-stationary, non-linear dynamics and (ii) separation of shared versus region-specific structure in a single framework. CTAE employs transformer encoders and decoders to capture long-range neural dynamics and explicitly partitions each region's latent space into orthogonal shared and private subspaces. We demonstrate the effectiveness of CTAE on two high-density electrophysiology datasets with simultaneous recordings from multiple regions, one from motor cortical areas and the other from sensory areas. CTAE extracts meaningful representations that better decode behavioral variables compared to existing approaches.

cs.LG↗

Joint Hierarchical Representation Learning of Samples and Features via Informed Tree-Wasserstein Distance

High-dimensional data often exhibit hierarchical structures in both modes: samples and features. Yet, most existing approaches for hierarchical representation learning consider only one mode at a time. In this work, we propose an unsupervised method for jointly learning hierarchical representations of samples and features via Tree-Wasserstein Distance (TWD). Our method alternates between the two data modes. It first constructs a tree for one mode, then computes a TWD for the other mode based on that tree, and finally uses the resulting TWD to build the second mode's tree. By repeatedly alternating through these steps, the method gradually refines both trees and the corresponding TWDs, capturing meaningful hierarchical representations of the data. We provide a theoretical analysis showing that our method converges. We show that our method can be integrated into hyperbolic graph convolutional networks as a pre-processing technique, improving performance in link prediction and node classification tasks. In addition, our method outperforms baselines in sparse approximation and unsupervised Wasserstein distance learning tasks on word-document and single-cell RNA-sequencing datasets.

cs.LG↗

Revisiting Meta-Learning with Noisy Labels: Reweighting Dynamics and Theoretical Guarantees

Learning with noisy labels remains challenging because over-parameterized networks memorize corrupted supervision. Meta-learning-based sample reweighting mitigates this by using a small clean subset to guide training, yet its behavior and training dynamics lack theoretical understanding. We provide a rigorous theoretical analysis of meta-reweighting under label noise and show that its training trajectory unfolds in three phases: (i) an alignment phase that amplifies examples consistent with a clean subset and suppresses conflicting ones; (ii) a filtering phase driving noisy example weights toward zero until the clean subset loss plateaus; and (iii) a post-filtering phase in which noise filtration becomes perturbation-sensitive. The mechanism is a similarity-weighted coupling between training and clean subset signals together with clean subset training loss contraction; in the post-filtering regime where the clean-subset loss is sufficiently small, the coupling term vanishes and meta-reweighting loses discriminatory power. Guided by this analysis, we propose a lightweight surrogate for meta-reweighting that integrates mean-centering, row shifting, and label-signed modulation, yielding more stable performance while avoiding expensive bi-level optimization. Across synthetic and real noisy-label benchmarks, our method consistently outperforms strong reweighting/selection baselines.

cs.LG↗

RnGCam: High-speed video from rolling & global shutter measurements

Compressive video capture encodes a short high-speed video into a single measurement using a low-speed sensor, then computationally reconstructs the original video. Prior implementations rely on expensive hardware and are restricted to imaging sparse scenes with empty backgrounds. We propose RnGCam, a system that fuses measurements from low-speed consumer-grade rolling-shutter (RS) and global-shutter (GS) sensors into video at kHz frame rates. The RS sensor is combined with a pseudorandom optic, called a diffuser, which spatially multiplexes scene information. The GS sensor is coupled with a conventional lens. The RS-diffuser provides low spatial detail and high temporal detail, complementing the GS-lens system's high spatial detail and low temporal detail. We propose a reconstruction method using implicit neural representations (INR) to fuse the measurements into a high-speed video. Our INR method separately models the static and dynamic scene components, while explicitly regularizing dynamics. In simulation, we show that our approach significantly outperforms previous RS compressive video methods, as well as state-of-the-art frame interpolators. We validate our approach in a dual-camera hardware setup, which generates 230 frames of video at 4,800 frames per second for dense scenes, using hardware that costs $10\times$ less than previous compressive video systems.

eess.IV↗