Randal Burns

Ariel Lubonja, Pedro R. A. S. Bassi, Wenxuan Li et al.

Open challenges have become the de facto standard for comparative ranking of medical AI methods. Despite their importance, medical AI leaderboards exhibit three persistent limitations: (1) score gaps are rarely tested for statistical significance, so rank stability is unknown; (2) single averaged metrics are applied to every organ, hiding clinically important boundary errors; (3) performance across intersecting demographics is seldom reported, masking fairness and equity gaps. We introduce RankInsight, an open-source toolkit that seeks to address these limitations. RankInsight (1) computes pair-wise significance maps that show the nnU-Net family outperforms Vision-Language and MONAI submissions with high statistical certainty; (2) recomputes leaderboards with organ-appropriate metrics, reversing the order of the top four models when Dice is replaced by NSD for tubular structures; and (3) audits intersectional fairness, revealing that more than half of the MONAI-based entries have the largest gender-race discrepancy on our proprietary Johns Hopkins Hospital dataset. The RankInsight toolkit is publicly released and can be directly applied to past, ongoing, and future challenges. It enables organizers and participants to publish rankings that are statistically sound, clinically meaningful, and demographically fair.

Robert Underwood, Meghana Madhastha, Randal Burns et al.

Network Architecture Search and specifically Regularized Evolution is a common way to refine the structure of a deep learning model.However, little is known about how models empirically evolve over time which has design implications for designing caching policies, refining the search algorithm for particular applications, and other important use cases.In this work, we algorithmically analyze and quantitatively characterize the patterns of model evolution for a set of models from the Candle project and the Nasbench-201 search space.We show how the evolution of the model structure is influenced by the regularized evolution algorithm. We describe how evolutionary patterns appear in distributed settings and opportunities for caching and improved scheduling. Lastly, we describe the conditions that affect when particular model architectures rise and fall in popularity based on their frequency of acting as a donor in a sliding window.

Leyang Xue, Meghana Madhyastha, Randal Burns et al.

Foundation models are at the forefront of AI research, appealing for their ability to learn from vast datasets and cater to diverse tasks. Yet, their significant computational demands raise issues of environmental impact and the risk of centralized control in their development. We put forward a vision towards decentralized and sustainable foundation model training that leverages the collective compute of sparingly used connected edge AI devices. We present the rationale behind our vision, particularly in support of its sustainability benefit. We further outline a set of challenges that need to be addressed to turn this vision into reality.

Ariel Lubonja, Cencheng Shen, Carey Priebe et al.

New algorithms for embedding graphs have reduced the asymptotic complexity of finding low-dimensional representations. One-Hot Graph Encoder Embedding (GEE) uses a single, linear pass over edges and produces an embedding that converges asymptotically to the spectral embedding. The scaling and performance benefits of this approach have been limited by a serial implementation in an interpreted language. We refactor GEE into a parallel program in the Ligra graph engine that maps functions over the edges of the graph and uses lock-free atomic instrutions to prevent data races. On a graph with 1.8B edges, this results in a 500 times speedup over the original implementation and a 17 times speedup over a just-in-time compiled version.

Leyang Xue, Meghana Madhyastha, Myungjin Lee et al.

The ecosystem behind foundation model development today is highly centralized and limited to large-scale cloud data center operators: training foundation models is costly, needing immense compute resources. Decentralized foundation model training across edge devices, leveraging their spare compute, promises a democratized alternative. However, existing edge-training approaches fall short: they struggle to match cloud-based training performance, exhibit limited scalability with model size, exceed device memory capacity, and have prohibitive communication overhead. They also fail to satisfactorily handle device heterogeneity and dynamism. We introduce a new paradigm, Cleave, which finely partitions training operations through a novel selective hybrid tensor parallelism method. Together with a parameter server centric training framework, Cleave copes with device memory limits and avoids communication bottlenecks, thereby enabling efficient training of large models on par with the cloud. Further, with a cost optimization model to guide device selection and training workload distribution, Cleave effectively accounts for device heterogeneity and churn. Our evaluations show that Cleave matches cloud-based GPU training by scaling efficiently to larger models and thousands of devices, supporting up to 8x more devices than baseline edge-training approaches. It outperforms state-of-the-art edge training methods by up to a factor of 10 in per-batch training time and efficiently handles device failures, achieving at least 100x faster recovery than prior methods.

Brian Wheatman, Meghana Madhyastha, Randal Burns

Artificial intelligence workloads, especially transformer models, exhibit emergent sparsity in which computations perform selective sparse access to dense data. The workloads are inefficient on hardware designed for dense computations and do not map well onto sparse data representations. We build a vectorized and parallel matrix-multiplication system A X B = C that eliminates unnecessary computations and avoids branches based on a runtime evaluation of sparsity. We use a combination of dynamic code lookup to adapt to the specific sparsity encoded in the B matrix and preprocessing of sparsity maps of the A and B matrices to compute conditional branches once for the whole computation. For a wide range of sparsity, from 60% to 95% zeros, our implementation performs fewer instructions and increases performance when compared with Intel MKL's dense or sparse matrix multiply routines. Benefits can be as large as 2 times speedup and 4 times fewer instructions.

Ashwin De Silva, Rahul Ramesh, Lyle Ungar et al.

Learning is a process which can update decision rules, based on past experience, such that future performance improves. Traditionally, machine learning is often evaluated under the assumption that the future will be identical to the past in distribution or change adversarially. But these assumptions can be either too optimistic or pessimistic for many problems in the real world. Real world scenarios evolve over multiple spatiotemporal scales with partially predictable dynamics. Here we reformulate the learning problem to one that centers around this idea of dynamic futures that are partially learnable. We conjecture that certain sequences of tasks are not retrospectively learnable (in which the data distribution is fixed), but are prospectively learnable (in which distributions may be dynamic), suggesting that prospective learning is more difficult in kind than retrospective learning. We argue that prospective learning more accurately characterizes many real world problems that (1) currently stymie existing artificial intelligence solutions and/or (2) lack adequate explanations for how natural intelligences solve them. Thus, studying prospective learning will lead to deeper insights and solutions to currently vexing challenges in both natural and artificial intelligences.

Meghana Madhyastha, Kunal Lillaney, James Browne et al.

We present methods to serialize and deserialize tree ensembles that optimize inference latency when models are not already loaded into memory. This arises whenever models are larger than memory, but also systematically when models are deployed on low-resource devices, such as in the Internet of Things, or run as Web micro-services where resources are allocated on demand. Our packed serialized trees (PACSET) encode reference locality in the layout of a tree ensemble using principles from external memory algorithms. The layout interleaves correlated nodes across multiple trees, uses leaf cardinality to collocate the nodes on the most popular paths and is optimized for the I/O blocksize. The result is that each I/O yields a higher fraction of useful data, leading to a 2-6 times reduction in classification latency for interactive workloads.

Meghana Madhyastha, Percy Li, James Browne et al.

Geodesic distance is the shortest path between two points in a Riemannian manifold. Manifold learning algorithms, such as Isomap, seek to learn a manifold that preserves geodesic distances. However, such methods operate on the ambient dimensionality, and are therefore fragile to noise dimensions. We developed an unsupervised random forest method (URerF) to approximately learn geodesic distances in linear and nonlinear manifolds with noise. URerF operates on low-dimensional sparse linear combinations of features, rather than the full observed dimensionality. To choose the optimal split in a computationally efficient fashion, we developed a fast Bayesian Information Criterion statistic for Gaussian mixture models. We introduce geodesic precision-recall curves which quantify performance relative to the true latent manifold. Empirical results on simulated and real data demonstrate that URerF is robust to high-dimensional noise, where as other methods, such as Isomap, UMAP, and FLANN, quickly deteriorate in such settings. In particular, URerF is able to estimate geodesic distances on a real connectome dataset better than other approaches.

Joshua T. Vogelstein, Eric Bridgeford, Minh Tang et al.

To solve key biomedical problems, experimentalists now routinely measure millions or billions of features (dimensions) per sample, with the hope that data science techniques will be able to build accurate data-driven inferences. Because sample sizes are typically orders of magnitude smaller than the dimensionality of these data, valid inferences require finding a low-dimensional representation that preserves the discriminating information (e.g., whether the individual suffers from a particular disease). There is a lack of interpretable supervised dimensionality reduction methods that scale to millions of dimensions with strong statistical theoretical guarantees.We introduce an approach, XOX, to extending principal components analysis by incorporating class-conditional moment estimates into the low-dimensional projection. The simplest ver-sion, "Linear Optimal Low-rank" projection (LOL), incorporates the class-conditional means. We prove, and substantiate with both synthetic and real data benchmarks, that LOL and its generalizations in the XOX framework lead to improved data representations for subsequent classification, while maintaining computational efficiency and scalability. Using multiple brain imaging datasets consisting of >150 million features, and several genomics datasets with>500,000 features, LOL outperforms other scalable linear dimensionality reduction techniques in terms of accuracy, while only requiring a few minutes on a standard desktop computer.

Tyler M. Tomita, James Browne, Cencheng Shen et al.

Decision forests, including Random Forests and Gradient Boosting Trees, have recently demonstrated state-of-the-art performance in a variety of machine learning settings. Decision forests are typically ensembles of axis-aligned decision trees; that is, trees that split only along feature dimensions. In contrast, many recent extensions to decision forests are based on axis-oblique splits. Unfortunately, these extensions forfeit one or more of the favorable properties of decision forests based on axis-aligned splits, such as robustness to many noise dimensions, interpretability, or computational efficiency. We introduce yet another decision forest, called "Sparse Projection Oblique Randomer Forests" (SPORF). SPORF uses very sparse random projections, i.e., linear combinations of a small subset of features. SPORF significantly improves accuracy over existing state-of-the-art algorithms on a standard benchmark suite for classification with >100 problems of varying dimension, sample size, and number of classes. To illustrate how SPORF addresses the limitations of both axis-aligned and existing oblique decision forest methods, we conduct extensive simulated experiments. SPORF typically yields improved performance over existing decision forests, while mitigating computational efficiency and scalability and maintaining interpretability. SPORF can easily be incorporated into other ensemble methods such as boosting to obtain potentially similar gains.

William Gray Roncal, Dean M. Kleissas, Joshua T. Vogelstein et al.

Reconstructing a map of neuronal connectivity is a critical challenge in contemporary neuroscience. Recent advances in high-throughput serial section electron microscopy (EM) have produced massive 3D image volumes of nanoscale brain tissue for the first time. The resolution of EM allows for individual neurons and their synaptic connections to be directly observed. Recovering neuronal networks by manually tracing each neuronal process at this scale is unmanageable, and therefore researchers are developing automated image processing modules. Thus far, state-of-the-art algorithms focus only on the solution to a particular task (e.g., neuron segmentation or synapse identification). In this manuscript we present the first fully automated images-to-graphs pipeline (i.e., a pipeline that begins with an imaged volume of neural tissue and produces a brain graph without any human interaction). To evaluate overall performance and select the best parameters and methods, we also develop a metric to assess the quality of the output graphs. We evaluate a set of algorithms and parameters, searching possible operating points to identify the best available brain graph for our assessment metric. Finally, we deploy a reference end-to-end version of the pipeline on a large, publicly available data set. This provides a baseline result and framework for community analysis and future algorithm development and testing. All code and data derivatives have been made publicly available toward eventually unlocking new biofidelic computational primitives and understanding of neuropathologies.

Ayushi Sinha, William Gray Roncal, Narayanan Kasthuri et al.

Accurately estimating the wiring diagram of a brain, known as a connectome, at an ultrastructure level is an open research problem. Specifically, precisely tracking neural processes is difficult, especially across many image slices. Here, we propose a novel method to automatically identify and annotate small subcellular structures present in axons, known as axoplasmic reticula, through a 3D volume of high-resolution neural electron microscopy data. Our method produces high precision annotations, which can help improve automatic segmentation by using our results as seeds for segmentation, and as cues to aid segment merging.

Ayushi Sinha, William Gray Roncal, Narayanan Kasthuri et al.

In this paper, we present a new pipeline which automatically identifies and annotates axoplasmic reticula, which are small subcellular structures present only in axons. We run our algorithm on the Kasthuri11 dataset, which was color corrected using gradient-domain techniques to adjust contrast. We use a bilateral filter to smooth out the noise in this data while preserving edges, which highlights axoplasmic reticula. These axoplasmic reticula are then annotated using a morphological region growing algorithm. Additionally, we perform Laplacian sharpening on the bilaterally filtered data to enhance edges, and repeat the morphological region growing algorithm to annotate more axoplasmic reticula. We track our annotations through the slices to improve precision, and to create long objects to aid in segment merging. This method annotates axoplasmic reticula with high precision. Our algorithm can easily be adapted to annotate axoplasmic reticula in different sets of brain data by changing a few thresholds. The contribution of this work is the introduction of a straightforward and robust pipeline which annotates axoplasmic reticula with high precision, contributing towards advancements in automatic feature annotations in neural EM data.

William Gray Roncal, Michael Pekala, Verena Kaynig-Fittkau et al.

An open challenge problem at the forefront of modern neuroscience is to obtain a comprehensive mapping of the neural pathways that underlie human brain function; an enhanced understanding of the wiring diagram of the brain promises to lead to new breakthroughs in diagnosing and treating neurological disorders. Inferring brain structure from image data, such as that obtained via electron microscopy (EM), entails solving the problem of identifying biological structures in large data volumes. Synapses, which are a key communication structure in the brain, are particularly difficult to detect due to their small size and limited contrast. Prior work in automated synapse detection has relied upon time-intensive biological preparations (post-staining, isotropic slice thicknesses) in order to simplify the problem. This paper presents VESICLE, the first known approach designed for mammalian synapse detection in anisotropic, non-post-stained data. Our methods explicitly leverage biological context, and the results exceed existing synapse detection methods in terms of accuracy and scalability. We provide two different approaches - one a deep learning classifier (VESICLE-CNN) and one a lightweight Random Forest approach (VESICLE-RF) to offer alternatives in the performance-scalability space. Addressing this synapse detection challenge enables the analysis of high-throughput imaging data soon expected to reach petabytes of data, and provide tools for more rapid estimation of brain-graphs. Finally, to facilitate community efforts, we developed tools for large-scale object detection, and demonstrated this framework to find $\approx$ 50,000 synapses in 60,000 $μm ^3$ (220 GB on disk) of electron microscopy data.