arXiv subjects

Luke McDermott

Publications and source records attributed to Luke McDermott.

12 recordsLinked to original sources

arXiv2026

Lifelong In-Context Learning with Transformers Requires Parametric Forms of Attention

Lifelong continual learning remains an obstacle on the path to human-like intelligence. Modern transformers show sparks of intelligence with in-context learning. The quadratic nature of attention, however, prohibits transformers from performing this process on arbitrarily long sequences. In this work, we argue that extending in-context learning to lifelong settings is a practical solution for continual learning in AI agents. In particular, we argue that \emph{parametric forms of attention} are needed to understand a lifetime of context with transformers on a fixed hardware budget. These attention mechanisms learn the relationship between keys and their associated values at test-time with parametric regression. Our generalization of parametric approaches (linear attention, state-space models, fast weight programmers, and test-time training layers) contrasts with nonparametric counterparts like softmax attention. They replace the ever-growing key-value cache with an online-trainable neural network, maintaining a constant memory footprint. We highlight how parametric attention currently fall short of lifelong learning due to limited memory capacity or costly online updates. To address these issues, we pose a set of open questions with novel insights to guide the field toward long-horizon agents.

cs.LG↗

arXiv2025

Identifying and Characterizing Very Low Mass Spectral Blend Binaries with Machine Learning Methods

We present an approach to identifying and characterizing unresolved, very low mass spectral blend binaries composed of late-M, L, and T dwarfs using machine learning methodologies. We generated and evaluated a series of hierarchical random forest models to distinguish spectral blends from single very low-mass dwarfs, and to classify their primary and secondary components. Models were trained on a sample of single and synthesized binary templates generated from empirical spectra. We explored various aspects of the design of our models, and find that models trained on a full range of single and binary combinations have the best performance for identification and component classification. These models achieve binary identification recall and precision of $\gtrsim$85%, median component classification errors of $\lesssim$0.1 subtypes, and systematic classification uncertainties of $\lesssim$1 subtype, outperforming index-based methods in terms of fidelity, range, and speed. Optimal performance is achieved for binaries composed of L and T dwarf primaries and late-L and T dwarf secondaries. When applied to the spectra of previously confirmed very low-mass binaries, model performance is degraded due to the prevalence of systems with similar component types, but remains high in the optimal performance range. We propose potential improvements to these models, which can be used to explore binary populations among the thousands to millions of very low-mass stars and brown dwarfs anticipated with large-scale spectral surveys such as SPHEREx and Euclid.

astro-ph.SR↗

arXiv2025

LoLA: Low-Rank Linear Attention With Sparse Caching

The per-token cost of transformer inference scales with context length, preventing its application to lifelong in-context learning. Linear attention is an efficient alternative that maintains a constant memory footprint, even on infinite context lengths. While this is a potential candidate for lifelong learning, it falls short in memory capacity. In this paper, we propose LoLA, a training-free augmentation to linear attention that boosts associative recall. LoLA distributes past key-value pairs from context into three memory systems: (i) recent pairs in a local sliding window cache; (ii) difficult-to-memorize pairs in a sparse, global cache; and (iii) generic pairs in the recurrent hidden state of linear attention. We show through ablations that our self-recall error metric is crucial to efficiently manage long-term associative memories. On pass-key retrieval tasks, LoLA improves the base model's performance from 0.6% to 97.4% accuracy. This is achieved with a 4.6x smaller cache than Llama-3.1 8B on 4K context length. LoLA also outperforms other 1B and 8B parameter subquadratic models on zero-shot commonsense reasoning tasks.

cs.CL↗

arXiv2025

Finding Stable Subnetworks at Initialization with Dataset Distillation

Recent works have shown that Dataset Distillation, the process for summarizing the training data, can be leveraged to accelerate the training of deep learning models. However, its impact on training dynamics, particularly in neural network pruning, remains largely unexplored. In our work, we use distilled data in the inner loop of iterative magnitude pruning to produce sparse, trainable subnetworks at initialization -- more commonly known as lottery tickets. While using 150x less training points, our algorithm matches the performance of traditional lottery ticket rewinding on ResNet-18 & CIFAR-10. Previous work highlights that lottery tickets can be found when the dense initialization is stable to SGD noise (i.e. training across different ordering of the data converges to the same minima). We extend this discovery, demonstrating that stable subnetworks can exist even within an unstable dense initialization. In our linear mode connectivity studies, we find that pruning with distilled data discards parameters that contribute to the sharpness of the loss landscape. Lastly, we show that by first generating a stable sparsity mask at initialization, we can find lottery tickets at significantly higher sparsities than traditional iterative magnitude pruning.

cs.LG↗

arXiv2025

Neural Architecture Codesign for Fast Physics Applications

We develop a pipeline to streamline neural architecture codesign for physics applications to reduce the need for ML expertise when designing models for novel tasks. Our method employs neural architecture search and network compression in a two-stage approach to discover hardware efficient models. This approach consists of a global search stage that explores a wide range of architectures while considering hardware constraints, followed by a local search stage that fine-tunes and compresses the most promising candidates. We exceed performance on various tasks and show further speedup through model compression techniques such as quantization-aware-training and neural network pruning. We synthesize the optimal models to high level synthesis code for FPGA deployment with the hls4ml library. Additionally, our hierarchical search space provides greater flexibility in optimization, which can easily extend to other tasks and domains. We demonstrate this with two case studies: Bragg peak finding in materials science and jet classification in high energy physics, achieving models with improved accuracy, smaller latencies, or reduced resource utilization relative to the baseline models.

cs.LG↗

arXiv2024

Embedding Compression for Efficient Re-Identification

Real world re-identfication (ReID) algorithms aim to map new observations of an object to previously recorded instances. These systems are often constrained by quantity and size of the stored embeddings. To combat this scaling problem, we attempt to shrink the size of these vectors by using a variety of compression techniques. In this paper, we benchmark quantization-aware-training along with three different dimension reduction methods: iterative structured pruning, slicing the embeddings at initialize, and using low rank embeddings. We find that ReID embeddings can be compressed by up to 96x with minimal drop in performance. This implies that modern re-identification paradigms do not fully leverage the high dimensional latent space, opening up further research to increase the capabilities of these systems.

cs.CV↗

arXiv2023

Neural Architecture Codesign for Fast Bragg Peak Analysis

We develop an automated pipeline to streamline neural architecture codesign for fast, real-time Bragg peak analysis in high-energy diffraction microscopy. Traditional approaches, notably pseudo-Voigt fitting, demand significant computational resources, prompting interest in deep learning models for more efficient solutions. Our method employs neural architecture search and AutoML to enhance these models, including hardware costs, leading to the discovery of more hardware-efficient neural architectures. Our results match the performance, while achieving a 13$\times$ reduction in bit operations compared to the previous state-of-the-art. We show further speedup through model compression techniques such as quantization-aware-training and neural network pruning. Additionally, our hierarchical search space provides greater flexibility in optimization, which can easily extend to other tasks and domains.

cs.LG↗

arXiv2023

Linear Mode Connectivity in Sparse Neural Networks

With the rise in interest of sparse neural networks, we study how neural network pruning with synthetic data leads to sparse networks with unique training properties. We find that distilled data, a synthetic summarization of the real data, paired with Iterative Magnitude Pruning (IMP) unveils a new class of sparse networks that are more stable to SGD noise on the real data, than either the dense model, or subnetworks found with real data in IMP. That is, synthetically chosen subnetworks often train to the same minima, or exhibit linear mode connectivity. We study this through linear interpolation, loss landscape visualizations, and measuring the diagonal of the hessian. While dataset distillation as a field is still young, we find that these properties lead to synthetic subnetworks matching the performance of traditional IMP with up to 150x less training points in settings where distilled data applies.

cs.LG↗

arXiv2023

UniCat: Crafting a Stronger Fusion Baseline for Multimodal Re-Identification

Multimodal Re-Identification (ReID) is a popular retrieval task that aims to re-identify objects across diverse data streams, prompting many researchers to integrate multiple modalities into a unified representation. While such fusion promises a holistic view, our investigations shed light on potential pitfalls. We uncover that prevailing late-fusion techniques often produce suboptimal latent representations when compared to methods that train modalities in isolation. We argue that this effect is largely due to the inadvertent relaxation of the training objectives on individual modalities when using fusion, what others have termed modality laziness. We present a nuanced point-of-view that this relaxation can lead to certain modalities failing to fully harness available task-relevant information, and yet, offers a protective veil to noisy modalities, preventing them from overfitting to task-irrelevant data. Our findings also show that unimodal concatenation (UniCat) and other late-fusion ensembling of unimodal backbones, when paired with best-known training techniques, exceed the current state-of-the-art performance across several multimodal ReID benchmarks. By unveiling the double-edged sword of "modality laziness", we motivate future research in balancing local modality strengths with global representations.

cs.CV↗

arXiv2023

GraFT: Gradual Fusion Transformer for Multimodal Re-Identification

Object Re-Identification (ReID) is pivotal in computer vision, witnessing an escalating demand for adept multimodal representation learning. Current models, although promising, reveal scalability limitations with increasing modalities as they rely heavily on late fusion, which postpones the integration of specific modality insights. Addressing this, we introduce the \textbf{Gradual Fusion Transformer (GraFT)} for multimodal ReID. At its core, GraFT employs learnable fusion tokens that guide self-attention across encoders, adeptly capturing both modality-specific and object-specific features. Further bolstering its efficacy, we introduce a novel training paradigm combined with an augmented triplet loss, optimizing the ReID feature embedding space. We demonstrate these enhancements through extensive ablation studies and show that GraFT consistently surpasses established multimodal ReID benchmarks. Additionally, aiming for deployment versatility, we've integrated neural network pruning into GraFT, offering a balance between model size and performance.

cs.CV↗

arXiv2023

A Generalization of Continuous Relaxation in Structured Pruning

Deep learning harnesses massive parallel floating-point processing to train and evaluate large neural networks. Trends indicate that deeper and larger neural networks with an increasing number of parameters achieve higher accuracy than smaller neural networks. This performance improvement, which often requires heavy compute for both training and evaluation, eventually needs to translate well to resource-constrained hardware for practical value. Structured pruning asserts that while large networks enable us to find solutions to complex computer vision problems, a smaller, computationally efficient sub-network can be derived from the large neural network that retains model accuracy but significantly improves computational efficiency. We generalize structured pruning with algorithms for network augmentation, pruning, sub-network collapse and removal. In addition, we demonstrate efficient and stable convergence up to 93% sparsity and 95% FLOPs reduction without loss of inference accuracy using with continuous relaxation matching or exceeding the state of the art for all structured pruning methods. The resulting CNN executes efficiently on GPU hardware without computationally expensive sparse matrix operations. We achieve this with routine automatable operations on classification and segmentation problems using CIFAR-10, ImageNet, and CityScapes datasets with the ResNet and U-NET network architectures.

cs.CV↗

arXiv2023

Distilled Pruning: Using Synthetic Data to Win the Lottery

This work introduces a novel approach to pruning deep learning models by using distilled data. Unlike conventional strategies which primarily focus on architectural or algorithmic optimization, our method reconsiders the role of data in these scenarios. Distilled datasets capture essential patterns from larger datasets, and we demonstrate how to leverage this capability to enable a computationally efficient pruning process. Our approach can find sparse, trainable subnetworks (a.k.a. Lottery Tickets) up to 5x faster than Iterative Magnitude Pruning at comparable sparsity on CIFAR-10. The experimental results highlight the potential of using distilled data for resource-efficient neural network pruning, model compression, and neural architecture search.

cs.LG↗