Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “representation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 289 records · Page 16
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

An Evaluation of Representation Learning Methods in Particle Physics Foundation Models

We present a systematic evaluation of representation learning objectives for particle physics within a unified framework. Our study employs a shared transformer-based particle-cloud encoder with standardized preprocessing, matched sampling, and a consistent evaluation protocol on a jet classification dataset. We compare contrastive (supervised and self-supervised), masked particle modeling, and generative reconstruction objectives under a common training regimen. In addition, we introduce targeted supervised architectural modifications that achieve state-of-the-art performance on benchmark evaluations. This controlled comparison isolates the contributions of the learning objective, highlights their respective strengths and limitations, and provides reproducible baselines. We position this work as a reference point for the future development of foundation models in particle physics, enabling more transparent and robust progress across the community.

Chen, Michael [Caltech]↗
DOE OSTI2025

Can Large Language Models Understand Intermediate Representations?

Intermediate Representations (IRs) are essential in compiler design and program analysis, yet their comprehension by Large Language Models (LLMs) remains underexplored. This paper presents a pioneering empirical study to investigate the capabilities of LLMs, including GPT-4, GPT-3, Gemma 2, LLaMA 3.1, and Code Llama, in understanding IRs. We analyze their performance across four tasks: Control Flow Graph (CFG) reconstruction, decompilation, code summarization, and execution reasoning. Our results indicate that while LLMs demonstrate competence in parsing IR syntax and recognizing high-level structures, they struggle with control flow reasoning, execution semantics, and loop handling. Specifically, they often misinterpret branching instructions, omit critical IR operations, and rely on heuristic-based reasoning, leading to errors in CFG reconstruction, IR decompilation, and execution reasoning. The study underscores the necessity for IR-specific enhancements in LLMs, recommending fine-tuning on structured IR datasets and integration of explicit control flow models to augment their comprehension and handling of IR-related tasks.

Jiang, Hailong↗
DOE OSTI2025

Contrasting Time-Frequency Representations for Unknown Waveform Detection

Identifying unseen electromagnetic waveforms is critical for many applications, like interference management, electronic warfare and spectrum management. Traditionally this is done using statistical methods for anomaly detection, which has evolved to deep learning models for identifying the unseen data, formally termed as open set recognition. Some prior methods use a generative model to emulate open set data, which face challenges in generating synthetic samples for open set while simultaneously selecting an optimal discriminator for accurate classification. To alleviate this issue, we propose a discriminative model that effectively combines time and frequency domain features of communication signals for accurate predictions. We further introduce a cosine similarity loss that makes the domain specific features unique to enhance the prediction rate. Additionally, our model avoids generic feature vectors by extracting class-specific features during training, resulting in improved class representation. The experiment results show that this combined feature approach with cosine loss outperforms single-domain models and improves accuracy by 10% over models without cosine loss.

99 - GENERAL AND MISCELLANEOUS↗
DOE OSTI2023

Python codes for the paper - "Trade-offs in the latent representation of microstructure evolution"

SAND2024-00946O Python codes used in "Trade-offs in the latent representation of microstructure evolution," a manuscript accepted for publication in Acta Materialia, are part of a repository. The code was developed to perform analysis of microstructure evolution. The repository consists of two main directories: models, which train and test models such as autoencoders and diffusion maps, and analysis, which analyzes microstructures. Source code is used to perform dimensionality reduction of microstructure data for analysis of its evolution in time. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Dingreville, Remi↗
DOE OSTI2020

Generation and representation of synthetic smart meter data

Advanced energy algorithms running at big-data scale will be necessary to identify, realize, and verify energy savings to meet government and utility goals of building energy efficiency. Any algorithm must be well characterized and validated before it is trusted to run at these scales. Smart meter data from real buildings will ultimately be required for the development, testing, and validation of these energy algorithms and processes. However, for initial development and testing, smart meter data are difficult to work with due to privacy restrictions, noise from unknown sources, data accessibility, and other concerns which can complicate algorithm development and validation. This paper describes a new methodology to generate synthetic smart meter data of electricity use in buildings using detailed building energy modeling, which aims to capture the variability and stochastics of real energy use in buildings. The methodology can create datasets tailored to represent specific scenarios with known truth and controllable amounts of synthetic noise. Knowledge of ground truth also allows the development and validation of enhanced processes which leverage building metadata, such as building type or size (floor area), in addition to smart meter data. The methodology described in this paper includes the key influencing factors of real-world building energy use including weather data, occupant-driven loads, building operation and maintenance practices, and special events. Data formats to support workflows leveraging both synthetic meter data and associated metadata are proposed and discussed. Finally, example use cases of the synthetic meter data are described to illustrate potential applications.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2021

The Stellar decomposition: A compact representation for simplicial complexes and beyond

Here, we introduce the Stellar decomposition, a model for efficient topological data structures over a broad range of simplicial and cell complexes. A Stellar decomposition of a complex is a collection of regions indexing the complex’s vertices and cells such that each region has sufficient information to locally reconstruct the star of its vertices, i.e., the cells incident in the region’s vertices. Stellar decompositions are general in that they can compactly represent and efficiently traverse arbitrary complexes with a manifold or non-manifold domain. They are scalable to complexes in high dimension and of large size, and they enable users to easily construct tailored application-dependent data structures using a fraction of the memory required by a corresponding global topological data structure on the complex. As a concrete realization of this model for spatially embedded complexes, we introduce the Stellar tree, which combines a nested spatial tree with a simple tuning parameter to control the number of vertices in a region. Stellar trees exploit the complex’s spatial locality by reordering vertex and cell indices according to the spatial decomposition and by compressing sequential ranges of indices. Stellar trees are competitive with state-of-the-art topological data structures for manifold simplicial complexes and offer significant improvements for cell complexes and non-manifold simplicial complexes. We conclude with a high-level description of several mesh processing and analysis applications that utilize Stellar trees to process large datasets.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Qualitative trend analysis based on a mixed-integer representation

Shape constrained spline fitting is a useful method to impose prior knowledge onto flexible semi-parametric models during parameter estimation. Most typically, the function shape is imposed through order restrictions on the regression coefficients. The intended shape is considered known or selected based on heuristic rules. In this study, we present a method to estimate the optimal set of order restrictions to segment a univariate data series into episodes with distinct shapes. This is also known as the qualitative trend analysis (QTA) problem. The obtained solution uses a trade-off between lack-of-fit and model complexity. Further, our practical implementation takes inspiration from the generalized order restricted information criterion (GORIC) for inequality-constrained model selection. From this, one learns (a) that QTA can be formulated as a mixed-integer quadratic program (MIQP) and (b) that the newly proposed mixed order restricted information criterion (MORIC) enables optimal segmentation. This is illustrated through didactic case studies.

42 ENGINEERING↗
DOE OSTI2023

Improving the Representation of Hydropower in Production Cost Models

One of the challenges of a high-renewable, low-carbon system is that grid operators will no longer be able to count on fossil-fired generation should wind miss its forecast or modeling errors create shortage issues, and this change has operational and reliability implications that affect variable renewable energy integration. Hydropower can help in that it can provide many of the services that have been traditionally provided by fossil generation. However, getting the most from hydropower can be challenging in that its capabilities vary with time and are affected by other water users' needs, two complexities that are not well represented in most of today's production cost models (PCMs). This work is an early step in addressing these gaps and making the most of hydropower's capabilities.

cost model↗
DOE OSTI2023

Aerodynamic Sensitivity of a Novel Data-Driven Airfoil Shape Representation Framework

We explore the aerodynamic implications of a novel data-driven separable shape tensor framework used to represent discrete airfoil shapes. In this study, we construct a data-driven parameter space defined by separable shape tensors and informed by tens of thousands of distinct airfoils. We use this design space to generate new airfoil designs to study parametric sensitivities with respect to various aerodynamic responses. We use a HAM2D RANS solver to approximate the lift, drag, and moment coefficients for the generated airfoils at two different angles-of-attack. We analyze the robustness and sensitivities of using the separable shape tensor design space by examining the coverage of the aerodynamic response space, uncovering low-dimensional polynomial ridge approximations, and computing various sensitivity metrics. The results show that the data-driven design space produce significant variation in target aerodynamic quantities and facilitate highly accurate approximations (R^2 > 0.96) of one- and two-dimensional structures in each aerodynamic response. This further reduces the effective dimension to enable simplified design and optimization tasks.

aerodynamics↗
DOE OSTI2026

Evaluation of a Reduced-Order Model for IBR Fault Response Representation via OEM Blackbox Models: Preprint

Driven by the need to capture the electromagnetic transients of transmission lines, inverter switching behavior, and detailed control systems, electromagnetic transient (EMT) studies have become increasingly important in industry, such as IBR interconnection study and fault study. However, original equipment manufacturer (OEM) inverter models typically include extensive parameters and proprietary settings that are unavailable to protection engineers. This paper introduces a data-driven, reduced-order model (ROM) developed as a PSCAD library component for use in EMT-based fault studies. The ROM replicates key OEM model behaviors without requiring detailed knowledge of control design or parameterization. The accompanying Python automation scripts streamline data generation, parameter fitting, and validation. The ROM's performance is demonstrated through comparison with both IEEE 2800-compliant and non-compliant OEM models in a real-world power system. Relay responses show nearly identical results, while simulation runtime is reduced by an average of 32.8\%, highlighting the ROM's practicality for protection engineers.

14 SOLAR ENERGY↗
DOE OSTI2024

The 2024 “Hacking Limnology” Workshop Series and Virtual Summit: Increasing Inclusion, Participation, and Representation in the Aquatic Sciences

The 4th Aquatic Ecosystem MOdeling Network—Junior (AEMON-J) Hacking Limnology Workshop and 5th Virtual Summit: Incorporating Data Science and Open Science in the Aquatic Sciences (DSOS) convened 15–19 July 2024. During the week, these joint communities engaged in activities at the intersection of big data, open science, modeling, remote sensing, and the aquatic sciences. The weeklong event, with over 100 aquatic science practitioners and enthusiasts, followed a similar structure to previous years, comprising three days of workshops followed by two days of the virtual summit.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Predicting Search Task Difficulty through a Discrete‐Time Action Log Representation on Spectrum Kernel

ABSTRACT Predicting perceived difficulty on a web search task is an open problem in the interactive information retrieval field. A common approach to tackle it, is through features obtained from full search sessions, which are then used to train classification models. In this poster we attempt to predict perceived task difficulty at different stages of the search process. To do so, we use the spectrum kernel for support vector machine (SVM) classification. Our preliminary results suggest that by using behavioral data from the first query segment, it is possible to provide timely classifications of whether a search task is perceived as hard or easy.

Gacitúa, Daniel↗
DOE OSTI2026

Physicochemical and Performance Characterization of Six Commercial Organic Solvent Nanofiltration Membranes

This work introduces a novel, gradient-free metamaterial design method based on Gaussian process regression to represent the density field of a unit cell. The dimension of the design space is determined by the covariance matrix dimension in the Gaussian process regression. We propose compressing this matrix using an autoencoder, enabling the decoder to generate the density field and effectively reduce the originally large design space to a lower-dimensional subspace. In this compressed space, we employ an active learning method, Bayesian Adaptive Direct Search (BADS), for efficient exploration of the design space. We demonstrate that for simple 2D designs aimed at maximizing unit cell stiffness, our method yields results comparable to those of standard topology optimization. Furthermore, we extend our approach to various mechanical problems, from linear elasticity to hyperelastic large deformation and elasto-plasticity under finite deformation, to 3D metamaterial design. This illustrates the method’s versatility and effectiveness across a range of applications.

Wu, Haoran↗
DOE OSTI2023

Efficient multi-scale representation of visual objects using a biologically plausible spike-latency code and winner-take-all inhibition

Deep neural networks have surpassed human performance in key visual challenges such as object recognition, but require a large amount of energy, computation, and memory. In contrast, spiking neural networks (SNNs) have the potential to improve both the efficiency and biological plausibility of object recognition systems. Here we present a SNN model that uses spike-latency coding and winner-take-all inhibition (WTA-I) to efficiently represent visual stimuli using multi-scale parallel processing. Mimicking neuronal response properties in early visual cortex, images were preprocessed with three different spatial frequency (SF) channels, before they were fed to a layer of spiking neurons whose synaptic weights were updated using spike-timing-dependent-plasticity. We investigate how the quality of the represented objects changes under different SF bands and WTA-I schemes. We demonstrate that a network of 200 spiking neurons tuned to three SFs can efficiently represent objects with as little as 15 spikes per neuron. Furthermore, studying how core object recognition may be implemented using biologically plausible learning rules in SNNs may not only further our understanding of the brain, but also lead to novel and efficient artificial vision systems.

59 BASIC BIOLOGICAL SCIENCES↗