arXiv subjects

Alexander DeLise

Publications and source records attributed to Alexander DeLise.

3 recordsLinked to original sources

arXiv2026

Active Learning for Conditional Generative Compressed Sensing

Generative compressed sensing uses the range of a pretrained generator as a nonlinear model for recovering structured signals from limited measurements. We study a conditional version of this problem for image recovery from subsampled Fourier measurements using prompt-conditioned generative models. Our framework separates two roles of conditioning: the prompt used to design the sampling distribution and the prompt used to define the recovery model. For ReLU and Lipschitz conditional generators, we prove stable recovery bounds showing that prompt-matched Christoffel sampling retains the same Christoffel complexity constant as existing near-optimal generative compressed sensing theory, while prompt mismatch incurs an explicit compatibility penalty. Experiments with Stable Diffusion show that prompts meaningfully reshape Christoffel sampling distributions and influence image recovery. Overall, our results suggest that prompts should be treated as design variables with distinct effects on sensing, approximation, and recovery.

cs.LG↗

arXiv2025

Optimal Linear Baseline Models for Scientific Machine Learning

Across scientific domains, a fundamental challenge is to characterize and compute the mappings from underlying physical processes to observed signals and measurements. While nonlinear neural networks have achieved considerable success, they remain theoretically opaque, which hinders adoption in contexts where interpretability is paramount. In contrast, linear neural networks serve as a simple yet effective foundation for gaining insight into these complex relationships. In this work, we develop a unified theoretical framework for analyzing linear encoder-decoder architectures through the lens of Bayes risk minimization for solving data-driven scientific machine learning problems. We derive closed-form, rank-constrained linear and affine linear optimal mappings for forward modeling and inverse recovery tasks. Our results generalize existing formulations by accommodating rank-deficiencies in data, forward operators, and measurement processes. We validate our theoretical results by conducting numerical experiments on datasets from simple biomedical imaging, financial factor analysis, and simulations involving nonlinear fluid dynamics via the shallow water equations. This work provides a robust baseline for understanding and benchmarking learned neural network models for scientific machine learning problems.

cs.LG↗

arXiv2025

Partitioned-Constraint QAOA (PC-QAOA): Structural State Preparation and Penalty Enforcement for Quantum Optimization

Constrained combinatorial optimization remains challenging for quantum algorithms because feasibility must be explicitly enforced, typically through penalty terms or problem-specific mixers. We introduce Partitioned-Constraint QAOA (PC-QAOA), which partitions constraints into those enforced structurally and those enforced energetically. Structural constraints are handled via feasible-state preparation and a Grover mixer that preserves feasibility, while the remaining constraints are enforced through penalties. We show that constraints with disjoint support can be prepared in parallel with little error accumulation. We identify broad classes of constraints (including cardinality, assignment, and flow conservation) that admit efficient structural enforcement, and introduce a variational gadget construction that extends this approach to arbitrary low-support constraints. Across 413 completed instances spanning multiple constraint families, PC-QAOA substantially improves feasibility and solution quality at shallow depth relative to penalty-based QAOA, demonstrating the value of partial structural enforcement.

quant-ph↗