SEARCH · Search NASA
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.
UWLi is a user interface for modifying Universal Workflow Language (UWL) files. UWL is a data format used to represent high fidelity scientific procedures in a generalized, field agnostic workflow format. See related publication: https://arxiv.org/pdf/2409.05899
You've found the official code repository for the paper "SysCaps: Language Interfaces for Simulation Surrogates of Complex Systems," presented at the Foundation Models for Science: Progress, Opportunities, and Challenges workshop at NeurIPS 2024. Our paper conjectures that interfaces (both text templates as well as conversational) makes interacting with simulation surrogate models for complex systems more intuitive and accessible for both non-experts and experts. "System captions", or SysCaps, are text-based descriptions of systems based on information contained in simulation metadata. Our paper's goal is to train multimodal regression models that take text inputs (SysCaps) and timeseries inputs (exogenous system conditions such as hourly weather) and regress timeseries simulation outputs (e.g. hourly building energy consumption). The experiments in our paper with building and wind farm simulators, which can be reproduced using this codebase, aim to help us understand whether a) accurate regression in this setting is possible and b) if so, how well can we do it. Paper: https://arxiv.org/abs/2405.19653
ChemReasoner - Catalyst Discovery via Large Language Model-driven Reasoning
In this study, we investigated strategies to address trust issues arising from errors in large language models (LLMs). The study examined the impact of confidence scores, system capability explanations, and user feedback on trust restoration post-error. 68 participants viewed the responses of an LLM to 20 general trivia questions, with an error introduced on the third trial. Each participant was presented with one mitigation strategy. Participants rated their overall trust in the model and the reliability of the answer. Results showed an immediate drop in trust after the error; however, there were no differences across the three strategies in trust recovery. All conditions had a logarithmic trend in trust recovery following error. Differences in overall trust were predicted by perceived reliability of the answer, suggesting that participants were evaluating results critically and using that to inform their trust in the model. Qualitative data supported this finding; participants expressed lasting distrust despite the LLM’s later accuracy. Results showcase the need to prioritize accuracy in LLM deployment, because early errors may irrevocably damage user trust calibration and later adoption.
Rule-based modeling is an approach that permits constructing reaction networks based on the specification of rules for molecular interactions and transformations. These rules can encompass details such as the interacting sub-molecular domains and the states and binding status of the involved components. Conceptually, fine-grained spatial information such as locations can also be provided. Through “wildcards” representing component states, entire families of molecule complexes sharing certain properties can be specified as patterns. This can significantly simplify the definition of models involving species with multiple components, multiple states, and multiple compartments. The systems biology markup language (SBML) Level 3 Multi Package Version 1 extends the SBML Level 3 Version 1 core with the “type” concept in the Species and Compartment classes. Therefore, reaction rules may contain species that can be patterns and exist in multiple locations. Multiple software tools such as Simmune and BioNetGen support this standard that thus also becomes a medium for exchanging rule-based models. This document provides the specification for Release 2 of Version 1 of the SBML Level 3 Multi package. No design changes have been made to the description of models between Release 1 and Release 2; changes are restricted to the correction of errata and the addition of clarifications.
We use model stitching to understand the internal representations of language models. Similar to vision models, we find that "more is better," and representations learned with more data and larger width can improve the performance of weaker models via stitching. We likewise find that certain architecture choices, using GeLU vs SoLU activation functions, influence the quality of learned representations. Finally, model stitching (as opposed to other model diagnostic methods, like mode connectivity) can localize the different generalization strategies of text classifiers under domain shift to certain hidden layers.
Discovering novel catalysts requires complex reasoning involving multiple chemical properties and resultant trade-offs, leading to a combinatorial growth in the search space. While large language models (LLM) have demonstrated novel capabilities for chemistry through complex instruction following capabilities and high quality reasoning, a goal-driven combinatorial search using LLMs has not been explored in detail. In this work, we present a Monte Carlo Tree Search-based approach that improves beyond state-of-the-art chain-of-thought prompting variants to augment scientific reasoning. We introduce two new reasoning datasets: 1) a curation of computational chemistry simulations, and 2) diverse questions written by catalysis researchers for reasoning about novel chemical conversion processes. We improve over the best baseline by 25.8\% and find that our approach can augment scientist's reasoning and discovery process with novel insights.\footnote{All resources will be publicly available upon publication.
Achieving interoperability involves reaching agreement on common interfaces and promoting interoperability through performance specifications. Effective integration practices are a simple, effective, and direct way of addressing the challenge. By deciding on an integration approach with well-defined interfaces and selecting systems and devices that conform to that approach and choosing integrators that have experience implementing the chosen approach, many problems can be avoided. No integration effort is seamless but by specifying interoperability requirements during the procurement process typical problems from ad hoc approaches can be avoided or reduced. By continuing this approach to its conclusion at the end of the procurement process, these same interoperability requirements may lead themselves to specified performance criteria that can be used to measure interoperability during implementation. The contractual language may contain performance incentives and penalties tied to these requirements as appropriate.
As the number of supported platforms for SNL software increases, so do the testing requirements. This increases the total time spent between when a developer submits code for testing, and when tests are completed. This in turn leads developers to hold off submitting code for testing, meaning that when code is ready for testing there's a lot more of it. This increases the likelihood of merge conflicts which the developer must resolve by hand -- because someone else touched the files near the lines the developer touched. Current text-based diff tools often have trouble resolving conflicts in these cases. Work in Europe and Japan has demonstrated that, using programming language aware diff tools (e.g., using the abstract syntax tree (AST) a compiler might generate) can reduce the manual labor necessary to resolve merge conflicts. These techniques can detect code blocks which have moved, as opposed than current text-based diff tools, which only detect insertions / deletions of text blocks. In this study, we evaluate one such tool, GumTree, and see how effective it is as a replacement for traditional text-based diff approaches.
Foreign disinformation campaigns undermine national security. Various supervised language modeling techniques in NLP can help to understand and dismantle these campaigns, but they rely heavily on large, labeled (often by humans) datasets. This work provides a solution to this problem in the form of an active learning (AL) framework, which is used to generate labeled datasets and leverage human input for detecting disinformation. The developed AL framework utilizes task adaptive pretraining to fully leverage the unlabeled data and boost the performance of the classifier used for labeling. A disinformation rhetoric metric was developed to measure the presence of common rhetorical techniques used in text that are meant to deceive, for both the classifier and human to use in the task of identifying disinformation. This metric was combined with an uncertainty criterion to create a hybrid acquisition method for AL, and this hybrid method was tested alongside other acquisition functions. A sophisticated and robust stopping strategy was developed to signal the AL process should terminate, saving human time from being wasted on iterations that would not significantly benefit classifier performance.
This final technical report summarizes the key accomplishments on the CALORIE project, a DOE Early Career award received by PI Henry Hoffmann at the University of Chicago. CALORIE’s main goal was to create principled methodologies, tools, and practices to help scientists and high-performance computing (HPC) operators maximize the performance and insights obtained from scientific computing applications in the face of exascale power constraints. The project accomplished these goals by completing the following objectives: • Designing a language for describing application goals (including constraints and objectives) and system capabilities. The application goals will include things like which simulations or simulations plus in situ analysis will be run together, what the power constraints are, and what requirements there are for in situ analysis (for example, a desired frame rate for visualization). The system capabilities include all components that can be adjusted to tradeoff power and performance. • Designing a runtime system that takes specified goals and capabilities and dynamically determines what capabilities to use to meet the goals. This runtime adapts to changes in goals, application behavior, or available capabilities to automatically maintain the goals despite unexpected disturbances. • Developing a foundational understanding of how power constraints affect the problem of scheduling applications in large-scale systems. This report provides an overview of the accomplishments related to each of these key objectives.
The goal of this project is to develop and evaluate a Machine Learning model using natural language processing techniques to categorize AD E-Log entries efficiently. AD E-Logs, which play a vital role in Fermilab's projects as a system to record and manage various data, often suffer from slow search processing and imprecise results. To address this issue, we preprocessed the data and employed the Doc2Vec model for training. The resulting model enables us to identify and retrieve the most relevant entries, thus enhancing user experience in accessing desired information.
The open-source Multiphysics Object-Oriented Simulation Environment (MOOSE) framework underpins most of the Nuclear Energy Advanced Modeling and Simulation (NEAMS) physics applications and coupling methods. Users interact with MOOSE in multiple ways to enable the construction of complex multiphysics models, including but not limited to compilation, input creation and syntax validation, meshing, solving the physics problem via MOOSE-based solvers and coupling, and output inspection. In FY23, numerous enhancements have been made to MOOSE to enhance the user experience. Reactor-oriented meshing capabilities in MOOSE have been expanded, an online tutorial was created and hosted on the MOOSE site, and a hands-on workshop was held for more than 100 users. The Language Server Protocol (LSP) capability has been implemented in MOOSE to better communicate correct syntax to the NEAMS Workbench user interface which is commonly used to validate input and submit jobs by users. Finally, several other user-facing improvements were made including the implementation of improved screen output and logging control, enhancements to the initial condition system, and enhancements of the MultiApp system commonly used for coupling. Collectively, these enhancements were driven by user needs and directly improve user ability to construct complex multiphysics models for nuclear reactors.
This document contains all relevant material generated during the authors' summer internship at NREL in 2024. This report shows how to improve energy efficiency of a few code samples by using low-precision data types combined with mixed-precision algorithms. The main applications considered here are (i) linear system solvers using mixed precision, and (ii) neural networks using mixed precision. This report also discusses how programming languages affect energy consumption of algorithms, energy metrics for a code and tools, and the available current software and hardware infrastructure.
A multidisciplinary team at Argonne National Laboratory evaluated the ability of large language models (LLMs) to identify geographic locations from open-source text and assessed post-processing measures to strengthen the reliability of those extractions in support of international nuclear safeguards. The study focused on addressing challenges such as toponym ambiguity, imprecise descriptions, and misinformation, which often undermine the accuracy of LLM-derived geospatial assessments. By integrating authoritative geospatial datasets, employing rigorous validation techniques, and leveraging human-in-the-loop processes, the project aimed to enhance the precision, transparency, and reproducibility of geospatial localization workflows. The findings demonstrate that while LLMs exhibit significant potential for accelerating geospatial analysis, their outputs require systematic grounding and verification to ensure reliability in high-stakes applications. This work contributes to the broader field of geospatial intelligence and supports strategic objectives of international organizations such as the International Atomic Energy Agency (IAEA) and the U.S. Department of Energy (DOE).
This work describes the improved workflow for evaluating open-source large language models (LLMs) for trustworthiness. The workflow facilitates the acquisition of LLMs, the generation of LLM responses, and the evaluation of the responses for their trustworthiness. As a use case, the workflow is employed to evaluate dense, quantized, and pruned Meta Llama3.1 LLMs for their truthfulness. The outcome of the project could set the stage for understanding and developing trustworthy models in the future projects.
The accelerated development and integration of large language model (LLM) agents have led researchers and developers to explore their effectiveness in cybersecurity, specifically with penetration testing (pentesting). Recent research efforts have attempted to use LLM agents to automate the process of pentesting because of the cost and time requirements that are required to perform a manual review. However, not all of the tools perform as expected. This paper reviews some of the newest and most popular autonomous pentesting frameworks, highlighting the capabilities and limitations of each one with the goal of providing the components needed to successfully and effectively build an autonomous pentesting agent in the future.
Large Language Models have proven adept at answering general knowledge questions. To make these generative AI tools useful to our mission customers for monitoring global threats, the data sciences team at Sandia is utilizing retrieval augmented generation (RAG) techniques to customize these models with local data. The local data we use consists of data such as research articles and patent abstracts that we've collected over the last several years using automated pipelines.