arXiv subjects

Eric Darve

Publications and source records attributed to Eric Darve.

At least 19 recordsLinked to original sources

arXiv2026

When Does Activation Steering Change What a Model Computes From?

Activation steering can reliably change an agent's output by modifying its internal activations. Yet arriving at the same answer need not involve the same computation: behavioral equivalence does not imply mechanistic equivalence. We test whether an activation edit changes the state used in subsequent computation or biases that computation toward the desired output. In a controlled state-tracking task, trace supervision creates an editable register: changing its internal value makes the model apply the next operation to the edited state. We next ask whether widely used mean activation steering admits a similar interpretation: does steering recreate the internal configuration the model naturally uses when performing a task? If so, transplanting that activation should be effective, and steering should remain effective at the scale of the natural source-to-target activation change. Neither prediction holds in the two selected model-task settings. Replacing the activation at one layer produces less than 2% of the target-answer margin gain from patching through all remaining layers, while natural-scale steering is similarly ineffective. The steering vectors have norms 77 and 26 times the median natural change in Qwen and Llama, producing 54% and 82% of the reference effect. Thus, a successful steering intervention need not reproduce the natural target activation at the intervention layer. More generally, an intervention should be interpreted as changing the computational state only when a later computation uses the edited value according to the semantics of that state.

cs.LG↗

arXiv2026

Discrete Wavelet Transform for Serial X-ray Crystallography Image Segmentation

Upcoming LCLS-II/II-HE operation at repetition rates approaching 1MHz demands on-detector data reduction to manage the resulting data volumes. We present a 2D discrete wavelet transform (DWT) pre-processing algorithm that segments background scatter from crystal diffraction in serial crystallography images, enabling early data analysis and, when combined with peak finding, lossy compression by transmitting only the identified diffraction peaks. The method zeroes the approximation (LL) coefficients of a multi-level Haar wavelet decomposition and reconstructs from detail subbands only, exploiting the natural separation of smooth background and sharp Bragg peaks in the wavelet domain. Evaluated on 100 simulated nanoBragg frames with known ground truth, the pipeline achieves $F1 \approx 0.96$ at four decomposition levels ($J = 4$), substantially outperforming the established peakfinder8 algorithm ($F1 \approx 0.37$) in both precision ($P \approx 1.00$ vs.\ $0.94$) and recall ($R \approx 0.92$ vs.\ $0.24$). A comparison of 12 wavelet families confirms that Haar is optimal for Bragg-peak detection due to its minimal filter support. Downstream crystallographic analysis performed on real ePix10kA data shows that CC* and $R_\mathrm{split}$ converge at $J = 4$ and track the unprocessed baseline through the practical resolution limit. Under added noise exceeding $\sim$50 ADU, the current pipeline's precision degrades significantly more than that of the pf8 algorithm, exposing a limitation of the proposed strategy. We also demonstrate an FPGA implementation of the DWT filters on an Alveo U200 at 200MHz, with a projected resource footprint compatible with integration into the upcoming ePixUHR firmware and a path to on-detector ASIC implementation in SparkPix detector family.

physics.ins-det↗

arXiv2026

Domain-Filtered Knowledge Graphs from Sparse Autoencoder Features

Sparse autoencoders (SAEs) extract millions of interpretable features from a language model, but flat feature inventories aren't very useful on their own. Domain concepts get mixed with generic and weakly grounded features, while related ideas are scattered across many units, and there's no way to understand relationships between features. We address this by first constructing a strict domain-specific concept universe from a large SAE inventory using contrastive activations and a multi-stage filtering process. Next, we build two aligned graph views on the filtered set: a co-occurrence graph for corpus-level conceptual structure, organized at multiple levels of granularity, and a transcoder-based mechanism graph that links source-layer and target-layer features through sparse latent pathways. Automated edge labeling then turns these graph views into readable knowledge graphs rather than unlabeled layouts. In a case study on a biology textbook, these graphs recover coherent chapter and subchapter-level structure, reveal concepts that bridge neighboring topics, and transform messy sentence-level activity containing thousands of features into compact, readable views that illustrate the model's local activity. Taken together, this reframes a flat SAE inventory as an internal knowledge graph that converts feature-level interpretability into a global map of model knowledge and enables audits of reasoning faithfulness.

cs.AI↗

arXiv2026

SpectraQuery: A Hybrid Retrieval-Augmented Conversational Assistant for Battery Science

Scientific reasoning increasingly requires linking structured experimental data with the unstructured literature that explains it, yet most large language model (LLM) assistants cannot reason jointly across these modalities. We introduce SpectraQuery, a hybrid natural-language query framework that integrates a relational Raman spectroscopy database with a vector-indexed scientific literature corpus using a Structured and Unstructured Query Language (SUQL)-inspired design. By combining semantic parsing with retrieval-augmented generation, SpectraQuery translates open-ended questions into coordinated SQL and literature retrieval operations, producing cited answers that unify numerical evidence with mechanistic explanation. Across SQL correctness, answer groundedness, retrieval effectiveness, and expert evaluation, SpectraQuery demonstrates strong performance: approximately 80 percent of generated SQL queries are fully correct, synthesized answers reach 93-97 percent groundedness with 10-15 retrieved passages, and battery scientists rate responses highly across accuracy, relevance, grounding, and clarity (4.1-4.6/5). These results show that hybrid retrieval architectures can meaningfully support scientific workflows by bridging data and discourse for high-volume experimental datasets.

cs.CL↗

arXiv2026

Necessary and Sufficient Conditions for the Existence of an LU Factorization for General Rank Deficient Matrices

We establish necessary and sufficient conditions for the existence of an LU factorization $A=LU$ for an arbitrary square matrix $A$, including singular and rank-deficient cases, without the use of row or column permutations. We prove that such a factorization exists if and only if the nullity of every leading principal submatrix is bounded by the sum of the nullities of the corresponding leading column and row blocks. While building upon the work of Okunev and Johnson, we present simpler, constructive proofs. Furthermore, we extend these results to characterize rank-revealing factorizations, providing explicit sparsity bounds for the factors $L$ and $U$. Finally, we derive analogous necessary and sufficient conditions for the existence of factorizations constrained to have unit lower or unit upper triangular factors.

math.NA↗

arXiv2025

Physically Interpretable Representation Learning with Gaussian Mixture Variational AutoEncoder (GM-VAE)

Extracting compact, physically interpretable representations from high-dimensional scientific data is a persistent challenge due to the complex, nonlinear structures inherent in physical systems. We propose a Gaussian Mixture Variational Autoencoder (GM-VAE) framework designed to address this by integrating an Expectation-Maximization (EM)-inspired training scheme with a novel spectral interpretability metric. Unlike conventional VAEs that jointly optimize reconstruction and clustering (often leading to training instability), our method utilizes a block-coordinate descent strategy, alternating between expectation and maximization steps. This approach stabilizes training and naturally aligns latent clusters with distinct physical regimes. To objectively evaluate the learned representations, we introduce a quantitative metric based on graph-Laplacian smoothness, which measures the coherence of physical quantities across the latent manifold. We demonstrate the efficacy of this framework on datasets of increasing complexity: surface reaction ODEs, Navier-Stokes wake flows, and experimental laser-induced combustion Schlieren images. The results show that our GM-VAE yields smooth, physically consistent manifolds and accurate regime clustering, offering a robust data-driven tool for interpreting turbulent and reactive flow systems.

cs.LG↗

arXiv2025

Multi-fidelity Batch Active Learning for Gaussian Process Classifiers

Many science and engineering problems rely on expensive computational simulations, where a multi-fidelity approach can accelerate the exploration of a parameter space. We study efficient allocation of a simulation budget using a Gaussian Process (GP) model in the binary simulation output case. This paper introduces Bernoulli Parameter Mutual Information (BPMI), a batch active learning algorithm for multi-fidelity GP classifiers. BPMI circumvents the intractability of calculating mutual information in the probability space by employing a first-order Taylor expansion of the link function. We evaluate BPMI against several baselines on two synthetic test cases and a complex, real-world application involving the simulation of a laser-ignited rocket combustor. In all experiments, BPMI demonstrates superior performance, achieving higher predictive accuracy for a fixed computational budget.

cs.LG↗

arXiv2025

Bi-fidelity Interpolative Decomposition for Multimodal Data

Multi-fidelity simulation is a widely used strategy to reduce the computational cost of many-query numerical simulation tasks such as uncertainty quantification, design space exploration, and design optimization. The reduced basis approach based on bi-fidelity interpolative decomposition is one such approach, which identifies a reduced basis, along with an interpolation rule in that basis, from low-fidelity samples to approximate the corresponding high-fidelity samples. However, as illustrated in the present study, when the model response is multi-modal and mode occupancy is stochastic, the assumptions underpinning this approach may not hold, thus leading to inaccurate estimates. We introduce the multi-modal interpolative decomposition method using bi-fidelity data, an extension tailored for this use case. Our work is motivated by a complex engineering application: a laser-ignited methane-oxygen rocket combustor evaluated over uncertain input parameters, exhibiting a bifurcation-like phenomenon in some regions of parameter space. Unlike the standard bi-fidelity interpolative decomposition approach, the proposed method can approximate a dataset of high-fidelity simulations for 16\% of the cost, while maintaining relatively high correlation (0.70--0.90) with parameter sensitivities.

math.NA↗

arXiv2025

Coincident Learning for Beam-based RF Station Fault Identification Using Phase Information at the SLAC Linac Coherent Light Source

Anomalies in radio-frequency (RF) stations can result in unplanned downtime and performance degradation in linear accelerators such as SLAC's Linac Coherent Light Source (LCLS). Detecting these anomalies is challenging due to the complexity of accelerator systems, high data volume, and scarcity of labeled fault data. Prior work identified faults using beam-based detection, combining RF amplitude and beam-position monitor data. Due to the simplicity of the RF amplitude data, classical methods are sufficient to identify faults, but the recall is constrained by the low-frequency and asynchronous characteristics of the data. In this work, we leverage high-frequency, time-synchronous RF phase data to enhance anomaly detection in the LCLS accelerator. Due to the complexity of phase data, classical methods fail, and we instead train deep neural networks within the Coincident Anomaly Detection (CoAD) framework. We find that applying CoAD to phase data detects nearly three times as many anomalies as when applied to amplitude data, while achieving broader coverage across RF stations. Furthermore, the rich structure of phase data enables us to cluster anomalies into distinct physical categories. Through the integration of auxiliary system status bits, we link clusters to specific fault signatures, providing additional granularity for uncovering the root cause of faults. We also investigate interpretability via Shapley values, confirming that the learned models focus on the most informative regions of the data and providing insight for cases where the model makes mistakes. This work demonstrates that phase-based anomaly detection for RF stations improves both diagnostic coverage and root cause analysis in accelerator systems and that deep neural networks are essential for effective analysis.

physics.acc-ph↗

arXiv2025

Physically Interpretable Representation and Controlled Generation for Turbulence Data

Computational Fluid Dynamics (CFD) plays a pivotal role in fluid mechanics, enabling precise simulations of fluid behavior through partial differential equations (PDEs). However, traditional CFD methods are resource-intensive, particularly for high-fidelity simulations of complex flows, which are further complicated by high dimensionality, inherent stochasticity, and limited data availability. This paper addresses these challenges by proposing a data-driven approach that leverages a Gaussian Mixture Variational Autoencoder (GMVAE) to encode high-dimensional scientific data into low-dimensional, physically meaningful representations. The GMVAE learns a structured latent space where data can be categorized based on physical properties such as the Reynolds number while maintaining global physical consistency. To assess the interpretability of the learned representations, we introduce a novel metric based on graph spectral theory, quantifying the smoothness of physical quantities along the latent manifold. We validate our approach using 2D Navier-Stokes simulations of flow past a cylinder over a range of Reynolds numbers. Our results demonstrate that the GMVAE provides improved clustering, meaningful latent structure, and robust generative capabilities compared to baseline dimensionality reduction methods. This framework offers a promising direction for data-driven turbulence modeling and broader applications in computational fluid dynamics and engineering systems.

cs.CE↗

arXiv2023

Factor Fitting, Rank Allocation, and Partitioning in Multilevel Low Rank Matrices

We consider multilevel low rank (MLR) matrices, defined as a row and column permutation of a sum of matrices, each one a block diagonal refinement of the previous one, with all blocks low rank given in factored form. MLR matrices extend low rank matrices but share many of their properties, such as the total storage required and complexity of matrix-vector multiplication. We address three problems that arise in fitting a given matrix by an MLR matrix in the Frobenius norm. The first problem is factor fitting, where we adjust the factors of the MLR matrix. The second is rank allocation, where we choose the ranks of the blocks in each level, subject to the total rank having a given value, which preserves the total storage needed for the MLR matrix. The final problem is to choose the hierarchical partition of rows and columns, along with the ranks and factors. This paper is accompanied by an open source package that implements the proposed methods.

stat.ML↗

arXiv2023

Resilient VAE: Unsupervised Anomaly Detection at the SLAC Linac Coherent Light Source

Significant advances in utilizing deep learning for anomaly detection have been made in recent years. However, these methods largely assume the existence of a normal training set (i.e., uncontaminated by anomalies) or even a completely labeled training set. In many complex engineering systems, such as particle accelerators, labels are sparse and expensive; in order to perform anomaly detection in these cases, we must drop these assumptions and utilize a completely unsupervised method. This paper introduces the Resilient Variational Autoencoder (ResVAE), a deep generative model specifically designed for anomaly detection. ResVAE exhibits resilience to anomalies present in the training data and provides feature-level anomaly attribution. During the training process, ResVAE learns the anomaly probability for each sample as well as each individual feature, utilizing these probabilities to effectively disregard anomalous examples in the training data. We apply our proposed method to detect anomalies in the accelerator status at the SLAC Linac Coherent Light Source (LCLS). By utilizing shot-to-shot data from the beam position monitoring system, we demonstrate the exceptional capability of ResVAE in identifying various types of anomalies that are visible in the accelerator.

physics.acc-ph↗

arXiv2023

A numerically stable communication-avoiding s-step GMRES algorithm

Krylov subspace methods are extensively used in scientific computing to solve large-scale linear systems. However, the performance of these iterative Krylov solvers on modern supercomputers is limited by expensive communication costs. The $s$-step strategy generates a series of $s$ Krylov vectors at a time to avoid communication. Asymptotically, the $s$-step approach can reduce communication latency by a factor of $s$. Unfortunately, due to finite-precision implementation, the step size has to be kept small for stability. In this work, we tackle the numerical instabilities encountered in the $s$-step GMRES algorithm. By choosing an appropriate polynomial basis and block orthogonalization schemes, we construct a communication avoiding $s$-step GMRES algorithm that automatically selects the optimal step size to ensure numerical stability. To further maximize communication savings, we introduce scaled Newton polynomials that can increase the step size $s$ to a few hundreds for many problems. An initial step size estimator is also developed to efficiently choose the optimal step size for stability. The guaranteed stability of the proposed algorithm is demonstrated using numerical experiments. In the process, we also evaluate how the choice of polynomial and preconditioning affects the stability limit of the algorithm. Finally, we show parallel scalability on more than 114,000 cores in a distributed-memory setting. Perfectly linear scaling has been observed in both strong and weak scaling studies with negligible communication costs.

math.NA↗

arXiv2023

Learning Reduced-Order Models for Cardiovascular Simulations with Graph Neural Networks

Reduced-order models based on physics are a popular choice in cardiovascular modeling due to their efficiency, but they may experience reduced accuracy when working with anatomies that contain numerous junctions or pathological conditions. We develop one-dimensional reduced-order models that simulate blood flow dynamics using a graph neural network trained on three-dimensional hemodynamic simulation data. Given the initial condition of the system, the network iteratively predicts the pressure and flow rate at the vessel centerline nodes. Our numerical results demonstrate the accuracy and generalizability of our method in physiological geometries comprising a variety of anatomies and boundary conditions. Our findings demonstrate that our approach can achieve errors below 2% and 3% for pressure and flow rate, respectively, provided there is adequate training data. As a result, our method exhibits superior performance compared to physics-based one-dimensional models, while maintaining high efficiency at inference time.

cs.LG↗

arXiv2023

Physics-based parameterized neural ordinary differential equations: prediction of laser ignition in a rocket combustor

In this work, we present a novel physics-based data-driven framework for reduced-order modeling of laser ignition in a model rocket combustor based on parameterized neural ordinary differential equations (PNODE). Deep neural networks are embedded as functions of high-dimensional parameters of laser ignition to predict various terms in a 0D flow model including the heat source function, pre-exponential factors, and activation energy. Using the governing equations of a 0D flow model, our PNODE needs only a limited number of training samples and predicts trajectories of various quantities such as temperature, pressure, and mass fractions of species while satisfying physical constraints. We validate our physics-based PNODE on solution snapshots of high-fidelity Computational Fluid Dynamics (CFD) simulations of laser-induced ignition in a prototype rocket combustor. We compare the performance of our physics-based PNODE with that of kernel ridge regression and fully connected neural networks. Our results show that our physics-based PNODE provides solutions with lower mean absolute errors of average temperature over time, thus improving the prediction of successful laser ignition with high-dimensional parameters.

cs.LG↗

arXiv2023

Coincident Learning for Unsupervised Anomaly Detection

Anomaly detection is an important task for complex systems (e.g., industrial facilities, manufacturing, large-scale science experiments), where failures in a sub-system can lead to low yield, faulty products, or even damage to components. While complex systems often have a wealth of data, labeled anomalies are typically rare (or even nonexistent) and expensive to acquire. Unsupervised approaches are therefore common and typically search for anomalies either by distance or density of examples in the input feature space (or some associated low-dimensional representation). This paper presents a novel approach called CoAD, which is specifically designed for multi-modal tasks and identifies anomalies based on \textit{coincident} behavior across two different slices of the feature space. We define an \textit{unsupervised} metric, $\hat{F}_\beta$, out of analogy to the supervised classification $F_\beta$ statistic. CoAD uses $\hat{F}_\beta$ to train an anomaly detection algorithm on \textit{unlabeled data}, based on the expectation that anomalous behavior in one feature slice is coincident with anomalous behavior in the other. The method is illustrated using a synthetic outlier data set and a MNIST-based image data set, and is compared to prior state-of-the-art on two real-world tasks: a metal milling data set and a data set from a particle accelerator.

cs.LG↗

arXiv2022

Probabilistic partition of unity networks for high-dimensional regression problems

We explore the probabilistic partition of unity network (PPOU-Net) model in the context of high-dimensional regression problems and propose a general framework focusing on adaptive dimensionality reduction. With the proposed framework, the target function is approximated by a mixture of experts model on a low-dimensional manifold, where each cluster is associated with a local fixed-degree polynomial. We present a training strategy that leverages the expectation maximization (EM) algorithm. During the training, we alternate between (i) applying gradient descent to update the DNN coefficients; and (ii) using closed-form formulae derived from the EM algorithm to update the mixture of experts model parameters. Under the probabilistic formulation, step (ii) admits the form of embarrassingly parallelizable weighted least-squares solves. The PPOU-Nets consistently outperform the baseline fully-connected neural networks of comparable sizes in numerical experiments of various data dimensions. We also explore the proposed model in applications of quantum computing, where the PPOU-Nets act as surrogate models for cost landscapes associated with variational quantum circuits.

cs.LG↗

arXiv2022

Beam-based RF Station Fault Identification at the SLAC Linac Coherent Light Source

Accelerators produce too many signals for a small operations team to monitor in real time. In addition, many of these signals are only interpretable by subject matter experts with years of experience. As a result, changes in accelerator performance can require time-intensive consultations with experts to identify the underlying problem. Herein, we focus on a particular anomaly detection task for radio-frequency (RF) stations at the Linac Coherent Light Source (LCLS). The existing RF station diagnostics are bandwidth limited, resulting in slow, unreliable signals. As a result, anomaly detection is currently a manual process. We propose a beam-based method, identifying changes in the accelerator status using shot-to-shot data from the beam position monitoring system; by comparing the beam-based anomalies to data from RF stations, we identify the source of the change. We find that our proposed method can be fully automated while identifying more events with fewer false positives than the RF station diagnostics alone. Our automated fault identification system has been used to create a new data set for investigating the interaction between the RF stations and accelerator performance.

physics.acc-ph↗