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 of computation. Understanding the functional principles of these networks is critical to developing ideal model architectures and optimization strategies. Previous studies typically only emphasize the network representation post-training, overlooking their evolution process throughout training. Here, we present Multiway Multislice PHATE (MM-PHATE), a novel method for visualizing the evolution of RNNs' hidden states. MM-PHATE is a graph-based embedding using structured kernels across the multiple dimensions spanned by RNNs: time, training epoch, and units. We demonstrate on various datasets that MM-PHATE uniquely preserves hidden representation community structure among units and identifies information processing and compression phases during training. The embedding allows users to look under the hood of RNNs across training and provides an intuitive and comprehensive strategy to understanding the network's internal dynamics and draw conclusions, e.g., on why and how one model outperforms another or how a specific architecture might impact an RNN's learning ability.

Comparing Graph Transformers via Positional Encodings

The distinguishing power of graph transformers is closely tied to the choice of positional encoding: features used to augment the base transformer with information about the graph. There are two primary types of positional encoding: absolute positional encodings (APEs) and relative positional encodings (RPEs). APEs assign features to each node and are given as input to the transformer. RPEs instead assign a feature to each pair of nodes, e.g., graph distance, and are used to augment the attention block. A priori, it is unclear which method is better for maximizing the power of the resulting graph transformer. In this paper, we aim to understand the relationship between these different types of positional encodings. Interestingly, we show that graph transformers using APEs and RPEs are equivalent in terms of distinguishing power. In particular, we demonstrate how to interchange APEs and RPEs while maintaining their distinguishing power in terms of graph transformers. Based on our theoretical results, we provide a study on several APEs and RPEs (including the resistance distance and the recently introduced stable and expressive positional encoding (SPE)) and compare their distinguishing power in terms of transformers. We believe our work will help navigate the huge number of choices of positional encoding and will provide guidance on the future design of positional encodings for graph transformers.

Deep and shallow data science for multi-scale optical neuroscience

Optical imaging of the brain has expanded dramatically in the past two decades. New optics, indicators, and experimental paradigms are now enabling in-vivo imaging from the synaptic to the cortex-wide scales. To match the resulting flood of data across scales, computational methods are continuously being developed to meet the need of extracting biologically relevant information. In this pursuit, challenges arise in some domains (e.g., SNR and resolution limits in micron-scale data) that require specialized algorithms. These algorithms can, for example, make use of state-of-the-art machine learning to maximally learn the details of a given scale to optimize the processing pipeline. In contrast, other methods, however, such as graph signal processing, seek to abstract away from some of the details that are scale-specific to provide solutions to specific sub-problems common across scales of neuroimaging. Here we discuss limitations and tradeoffs in algorithmic design with the goal of identifying how data quality and variability can hamper algorithm use and dissemination.

Learning Cartesian Product Graphs with Laplacian Constraints

Graph Laplacian learning, also known as network topology inference, is a problem of great interest to multiple communities. In Gaussian graphical models (GM), graph learning amounts to endowing covariance selection with the Laplacian structure. In graph signal processing (GSP), it is essential to infer the unobserved graph from the outputs of a filtering system. In this paper, we study the problem of learning Cartesian product graphs under Laplacian constraints. The Cartesian graph product is a natural way for modeling higher-order conditional dependencies and is also the key for generalizing GSP to multi-way tensors. We establish statistical consistency for the penalized maximum likelihood estimation (MLE) of a Cartesian product Laplacian, and propose an efficient algorithm to solve the problem. We also extend our method for efficient joint graph learning and imputation in the presence of structural missing values. Experiments on synthetic and real-world datasets demonstrate that our method is superior to previous GSP and GM methods.

Contextual Feature Selection with Conditional Stochastic Gates

We study the problem of contextual feature selection, where the goal is to learn a predictive function while identifying subsets of informative features conditioned on specific contexts. Towards this goal, we generalize the recently proposed stochastic gates (STG) Yamada et al. [2020] by modeling the probabilistic gates as conditional Bernoulli variables whose parameters are predicted based on the contextual variables. Our new scheme, termed conditional-STG (c-STG), comprises two networks: a hypernetwork that establishes the mapping between contextual variables and probabilistic feature selection parameters and a prediction network that maps the selected feature to the response variable. Training the two networks simultaneously ensures the comprehensive incorporation of context and feature selection within a unified model. We provide a theoretical analysis to examine several properties of the proposed framework. Importantly, our model leads to improved flexibility and adaptability of feature selection and, therefore, can better capture the nuances and variations in the data. We apply c-STG to simulated and real-world datasets, including healthcare, housing, and neuroscience, and demonstrate that it effectively selects contextually meaningful features, thereby enhancing predictive performance and interpretability.

Semi-Supervised Laplacian Learning on Stiefel Manifolds

Motivated by the need to address the degeneracy of canonical Laplace learning algorithms in low label rates, we propose to reformulate graph-based semi-supervised learning as a nonconvex generalization of a \emph{Trust-Region Subproblem} (TRS). This reformulation is motivated by the well-posedness of Laplacian eigenvectors in the limit of infinite unlabeled data. To solve this problem, we first show that a first-order condition implies the solution of a manifold alignment problem and that solutions to the classical \emph{Orthogonal Procrustes} problem can be used to efficiently find good classifiers that are amenable to further refinement. Next, we address the criticality of selecting supervised samples at low-label rates. We characterize informative samples with a novel measure of centrality derived from the principal eigenvectors of a certain submatrix of the graph Laplacian. We demonstrate that our framework achieves lower classification error compared to recent state-of-the-art and classical semi-supervised learning methods at extremely low, medium, and high label rates. Our code is available on github\footnote{anonymized for submission}.

Graph Laplacian Learning with Exponential Family Noise

A common challenge in applying graph machine learning methods is that the underlying graph of a system is often unknown. Although different graph inference methods have been proposed for continuous graph signals, inferring the graph structure underlying other types of data, such as discrete counts, is under-explored. In this paper, we generalize a graph signal processing (GSP) framework for learning a graph from smooth graph signals to the exponential family noise distribution to model various data types. We propose an alternating algorithm that estimates the graph Laplacian as well as the unobserved smooth representation from the noisy signals. We demonstrate in synthetic and real-world data that our new algorithm outperforms competing Laplacian estimation methods under noise model mismatch.

SiBBlInGS: Similarity-driven Building-Block Inference using Graphs across States

Interpretable methods for extracting meaningful building blocks (BBs) underlying multi-dimensional time series are vital for discovering valuable insights in complex systems. Existing techniques, however, encounter limitations that restrict their applicability to real-world systems, like reliance on orthogonality assumptions, inadequate incorporation of inter- and intra-state variability, and incapability to handle sessions of varying duration. Here, we present a framework for Similarity-driven Building Block Inference using Graphs across States (SiBBlInGS). SiBBlInGS employs a graph-based dictionary learning approach for BB discovery, simultaneously considers both inter- and intra-state relationships in the data, can extract non-orthogonal components, and allows for variations in session counts and duration across states. Additionally, SiBBlInGS allows for cross-state variations in BB structure and per-trial temporal variability, can identify state-specific vs state-invariant BBs, and offers both supervised and data-driven approaches for controlling the level of BB similarity between states. We demonstrate SiBBlInGS on synthetic and real-world data to highlight its ability to provide insights into the underlying mechanisms of complex phenomena and its applicability to data in various fields.

Hyperbolic Diffusion Embedding and Distance for Hierarchical Representation Learning

Finding meaningful representations and distances of hierarchical data is important in many fields. This paper presents a new method for hierarchical data embedding and distance. Our method relies on combining diffusion geometry, a central approach to manifold learning, and hyperbolic geometry. Specifically, using diffusion geometry, we build multi-scale densities on the data, aimed to reveal their hierarchical structure, and then embed them into a product of hyperbolic spaces. We show theoretically that our embedding and distance recover the underlying hierarchical structure. In addition, we demonstrate the efficacy of the proposed method and its advantages compared to existing methods on graph embedding benchmarks and hierarchical datasets.

DiSC: Differential Spectral Clustering of Features

Selecting subsets of features that differentiate between two conditions is a key task in a broad range of scientific domains. In many applications, the features of interest form clusters with similar effects on the data at hand. To recover such clusters we develop DiSC, a data-driven approach for detecting groups of features that differentiate between conditions. For each condition, we construct a graph whose nodes correspond to the features and whose weights are functions of the similarity between them for that condition. We then apply a spectral approach to compute subsets of nodes whose connectivity differs significantly between the condition-specific feature graphs. On the theoretical front, we analyze our approach with a toy example based on the stochastic block model. We evaluate DiSC on a variety of datasets, including MNIST, hyperspectral imaging, simulated scRNA-seq and task fMRI, and demonstrate that DiSC uncovers features that better differentiate between conditions compared to competing methods.