Search NASA⌕ Search

SEARCH · Search NASA

Results for “Small Language Model”

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 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

ESAC (EQ-SANS Assisting Chatbot): Application of large language models and retrieval-augmented generation for enhanced user experience at EQ-SANS

Neutron scattering experiments have played vital roles in exploring materials properties in the past decades. While user interfaces have been improved over time, neutron scattering experiments still require specific knowledge or training by an expert due to the complexity of such advanced instrumentation and the limited number of experiments each person may perform each year. This paper introduces an innovative chatbot application that leverages Large Language Models(LLM) and Retrieval-Augmented Generation (RAG) technologies to significantly enhance the user experience at the EQ-SANS, a small-angle neutron scattering instrument at the Spallation Neutron Source of Oak Ridge National Laboratory. Through a user-centric design approach, the EQ-SANS Assisting Chatbot (ESAC) serves as an interactive reference for users, thereby facilitating the use of the instrument by visiting scientists. By bridging the gap between the users of EQ-SANS and the control systems required to perform their experiments, the ESAC sets a new standard for interactive learning and support for the scientific community using large-scale scientific facilities.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Semantic Stealth: Crafting Covert Adversarial Patches for Sentiment Classifiers Using Large Language Models

Deep learning models have been shown to be vulnerable to adversarial attacks, in which perturbations to their inputs cause the model to produce incorrect predictions. As opposed to adversarial attacks in computer vision, where small changes introduced to pixel values can drastically alter a model's output while remaining imperceptible to humans, text-based attacks are difficult to conceal due to the discrete nature of tokens. Consequently, unconstrained gradient-based attacks often produce adversarial examples that lack semantic meaning, rendering them detectable through visual inspection or perplexity filters. In contrast to methods that rely on gradient-based optimization in the embedding space, we propose an approach that leverages a Large Language Model's ability to generate grammatically correct and semantically meaningful text to craft adversarial patches that seamlessly blend in with the original input text. These patches can be used to alter the behavior of a target model, such as a text classifier. Since our approach does not rely on gradient backpropagation, it only requires access to the target model's confidence scores, making it a grey-box attack. We demonstrate the feasibility of our approach using open-source LLMs, including Intel's Neural Chat, Llama2, and Mistral-Instruct, to generate adversarial patches capable of altering the predictions of a distilBERT model fine-tuned on the IMDB reviews dataset for sentiment classification.

Roa Carvajal, Maria↗
DOE OSTI2024

Evaluation of the Self Retrieval Augmented Generation Technique on Common Security Advisory Framework Data

This small experimental report evaluates a variation of Retrieval Augmented Generation (RAG), called Self-RAG. This method uses a generative language model that incorporates retrieved facts into its generation and is explicitly trained to be able to determine whether retrieved information is enough to answer the input query, with a user-defined threshold for confidence. We performed an experiment using data from the publicly available CISA Common Security Advisory Framework (CSAF) repository (https://github.com/cisagov/CSAF) as the database of facts to be used in retrieval. Qualitative results from the experiment demonstrate that the Self-RAG method has some ability to provide reasonable answers to queries that are in the dataset and will often ignore irrelevant information when asked outside of domain questions (e.g., general facts). In settings with deliberately confusing questions (the question is within domain, but asks about a fabricated advisory), it was able to refuse 40% of the time without further adjustments to the original framework. While this performance is not sufficient for current practical use, further improvements to data formatting, disambiguating results, and leveraging threshold values could improve performance significantly. However, evaluating this will require more extensive evaluations on larger datasets and potentially better models.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Do Programmers Prefer Predictable Expressions in Code?

Source code is a form of human communication, albeit one where the information shared between the programmers reading and writing the code is constrained by the requirement that the code executes correctly. Programming languages are more syntactically constrained than natural languages, but they are also very expressive, allowing a great many different ways to express even very simple computations. Still, code written by developers is highly predictable, and many programming tools have taken advantage of this phenomenon, relying on language model surprisal as a guiding mechanism. Additionally, while surprisal has been validated as a measure of cognitive load in natural language, its relation to human cognitive processes in code is still poorly understood. In this paper, we explore the relationship between surprisal and programmer preference at a small granularity—do programmers prefer more predictable expressions in code? Using meaning-preserving transformations, we produce equivalent alternatives to developer-written code expressions and run a corpus study on Java and Python projects. In general, language models rate the code expressions developers choose to write as more predictable than these transformed alternatives. Then, we perform two human subject studies asking participants to choose between two equivalent snippets of Java code with different surprisal scores (one original and transformed). We find that programmers do prefer more predictable variants, and that stronger language models like the transformer align more often and more consistently with these preferences.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1991

Simulation-based instruction of technical skills

A rapid intelligent tutoring development system (RAPIDS) was developed to facilitate the production of interactive, real-time graphical device models for use in instructing the operation and maintenance of complex systems. The tools allowed subject matter experts to produce device models by creating instances of previously defined objects and positioning them in the emerging device model. These simulation authoring functions, as well as those associated with demonstrating procedures and functional effects on the completed model, required no previous programming experience or use of frame-based instructional languages. Three large simulations were developed in RAPIDS, each involving more than a dozen screen-sized sections. Seven small, single-view applications were developed to explore the range of applicability. Three workshops were conducted to train others in the use of the authoring tools. Participants learned to employ the authoring tools in three to four days and were able to produce small working device models on the fifth day.

Towne, Douglas M.↗
NASA NTRSDate not supplied

Biological Research and Space Health Enabled by Machine Learning to Support Deep Space Missions

A key science goal of the NASA “Moon to Mars” campaign is to understand how biology responds to the Lunar, Martian, and deep space environments in order to advance fundamental knowledge, reduce risk, and support safe, productive human space missions. Through the powerful emerging computer science approaches of artificial intelligence (AI) and machine learning (ML), a paradigm shift has begun in biomedical science and engineered astronaut health systems, to enable Earth-independence and autonomy of mission operations. We present a decadal view of AI/ML architecture to support deep space mission goals, developed in concert with leaders in the field. We describe current AI/ML methods to support 1) fundamental biology, 2) in situ analytics, 3) high performance computing hardware, 4) automated science, 5) self-driving labs, 6) remote data management, 7) integrated real-time mission biomonitoring, and 8) a Precision Space Health system. Cutting-edge AI/ML approaches that can be integrated to support these domains include active learning, explainable AI, adaptive learning, causal inference, knowledge graphs, federated learning, transfer learning, and large language models. Finally, we present results from several current ML projects that are underway in the field to address key challenges of small sample n, high feature count, heterogeneity, and sparse data. These include 1) connecting omics data to phenotypic data using an ensemble model to infer causality of spaceflight rodent liver health disruption, 2) usage of explainable ML to interrogate the muscular underpinnings of spaceflight muscle atrophy, 3) ML models analyzing and determining directed acyclic graphs of human space health risk leveraging rodent bone datasets, 4) usage of large pre-trained models connecting biomedical knowledgebases with small spaceflight datasets to understand gene-to-gene interaction networks, and 5) a suite of benchmarked open science datasets (spaceflight mouse liver; radiation DNA damage) enabling programmers to identify the best ML algorithms to answer space biological science questions.

space biology↗
NASA NTRSDate not supplied

Biological Research and Space Health Enabled by Machine Learning to Support Deep Space Missions

A key science goal of the NASA “Moon to Mars” campaign is to understand how biology responds to the Lunar, Martian, and deep space environments in order to advance fundamental knowledge, reduce risk, and support safe, productive human space missions. Through the powerful emerging computer science approaches of artificial intelligence (AI) and machine learning (ML), a paradigm shift has begun in biomedical science and engineered astronaut health systems, to enable Earth-independence and autonomy of mission operations. We present a decadal view of AI/ML architecture to support deep space mission goals, developed in concert with leaders in the field. We describe current AI/ML methods to support 1) fundamental biology, 2) in situ analytics, 3) high performance computing hardware, 4) automated science, 5) self-driving labs, 6) remote data management, 7) integrated real-time mission biomonitoring, and 8) a Precision Space Health system. Cutting-edge AI/ML approaches that can be integrated to support these domains include active learning, explainable AI, adaptive learning, causal inference, knowledge graphs, federated learning, transfer learning, and large language models. Finally, we present results from several current ML projects that are underway in the field to address key challenges of small sample n, high feature count, heterogeneity, and sparse data. These include 1) connecting omics data to phenotypic data using an ensemble model to infer causality of spaceflight rodent liver health disruption, 2) usage of explainable ML to interrogate the muscular underpinnings of spaceflight muscle atrophy, 3) ML models analyzing and determining directed acyclic graphs of human space health risk leveraging rodent bone datasets, 4) usage of large pre-trained models connecting biomedical knowledgebases with small spaceflight datasets to understand gene-to-gene interaction networks, and 5) a suite of benchmarked open science datasets (spaceflight mouse liver; radiation DNA damage) enabling programmers to identify the best ML algorithms to answer space biological science questions.

space biology↗
DOE OSTI2026

MTRE

Multi-Token Reliability Estimation (MTRE) is a lightweight, white-box hallucination detector for vision-language models. Instead of using only the first output token, MTRE aggregates logits from the first ~10 tokens and feeds them to a small attention-based reliability head; per-token scores are combined via a sequential log-likelihood-ratio test with early-stopping, and an MTRE-t variant calibrates thresholds via cross-fitting. MTRE reports average gains of +9.4% Accuracy and +14.8% AUROC over common baselines across MAD-Bench, MM-SafetyBench, MathVista, and arithmetic/counting tasks, while adding ~4.3M params and ~1% inference overhead (~26 MB VRAM, ~0.94 ms per detection). Key limitation: requires access to early token logits and is evaluated on a handful of open-source 7B VLMs.

Bhattarai, Manish [Los Alamos National Labs]↗
NASA NTRSDate not supplied

Biological Research and Space Health Enabled by Machine Learning to Support Deep Space Missions

A key science goal of the NASA “Moon to Mars” campaign is to understand how biology responds to the Lunar, Martian, and deep space environments in order to advance fundamental knowledge and support human space missions. Through artificial intelligence (AI) and machine learning (ML), a paradigm shift has begun in space biosciences and engineered astronaut health systems, to enable Earth-independence and mission operations autonomy. We describe current AI/ML methods to support 1) fundamental biology, 2) in situ analytics, 3) high performance computing, 4) automated science, 5) self-driving labs, 6) remote data management, 7) integrated mission biomonitoring, and 8) a Precision Space Health system. AI/ML approaches that can be integrated to support these domains include active learning, explainable AI, adaptive learning, causal inference, knowledge graphs, federated learning, transfer learning, and large language models. Finally, we present results from several current ML projects that are underway in the space biology field to address key challenges of small sample n, high feature count, heterogeneity, and sparse data. These include 1) connecting omics to phenotypic data using an ensemble model to infer causality of rodent liver health disruption, 2) usage of explainable ML to interrogate muscular underpinnings of muscle atrophy, 3) ML models analyzing and determining directed acyclic graphs of human health risk leveraging rodent bone datasets, 4) usage of large pre-trained models connecting biomedical knowledgebases with small spaceflight datasets to understand gene-to-gene interactions, and 5) a suite of benchmarked open science datasets enabling programmers to identify best algorithms to answer space biology questions.

space biology↗
NASA NTRS2011

Theoretical Near-IR Spectra for Surface Abundance Studies of Massive Stars

We present initial results of a study of abundance and mass loss properties of O-type stars based on theoretical near-IR spectra computed with state-of-the-art stellar atmosphere models. The James Webb Space Telescope (JWST) will be a powerful tool to obtain high signal-to-noise ratio near-IR (1-5 micron) spectra of massive stars in different environments of local galaxies. Our goal is to analyze model near-IR spectra corresponding to those expected from NIRspec on JWST in order to map the wind properties and surface composition across the parameter range of 0 stars and to determine projected rotational velocities. As a massive star evolves, internal coupling, related mixing, and mass loss impact its intrinsic rotation rate. These three parameters form an intricate loop, where enhanced rotation leads to more mixing which in turn changes the mass loss rate, the latter thus affecting the rotation rate. Since the effects of rotation are expected to be much more pronounced at low metallicity, we pay special attention to models for massive stars in the the Small Magellanic Cloud. This galaxy provides a unique opportunity to probe stellar evolution, and the feedback of massive stars on galactic evol.ution in conditions similar to the epoch of maximal star formation. Plain-Language Abstract: We present initial results of a study of abundance and mass loss properties of massive stars based on theoretical near-infrared (1-5 micron) spectra computed with state-of-the-art stellar atmosphere models. This study is to prepare for observations by the James Webb Space Telescope.

Sonneborn, George↗
NASA NTRS1984

A real-time, portable, microcomputer-based jet engine simulator

Modern piloted flight simulators require detailed models of many aircraft components, such as the airframe, propulsion system, flight deck controls and instrumentation, as well as motion drive and visual display systems. The amount of computing power necessary to implement these systems can exceed that offered by dedicated mainframe computers. One approach to this problem is through the use of distributed computing, where parts of the simulation are assigned to computing subsystems, such as microcomputers. One such subsystem, such as microcomputers. One such subsystem, a real-time, portable, microcomputer-based jet engine simulator, is described in this paper. The simulator will be used at the NASA Ames Vertical Motion Simulator facility to perform calculations previously done on the facility's mainframe computer. The mainframe will continue to do all other system calculations and will interface to the engine simulator through analog I/0. The engine simulator hardware includes a 16-bit microcomputer and floating-point coprocessor. There is an 8 channel analog input board and an 8 channel analog output board. A model of a small turboshaft engine/control is coded in floating-point FORTRAN. The FORTRAN code and a data monitoring program run under the control of an assembly language real-time executive. The monitoring program allows the user to isplay and/or modify simulator variables on-line through a data terminal. A dual disk drive system is used for mass storage of programs and data. The CP/M-86 operating system provides file management and overall system control. The frame time for the simulator is 30 milliseconds, which includes all analog I/0 operations.

Blech, R. A.↗
NASA NTRS1985

Massively parallel processor

A brief description is given of the Massively Parallel Processor (MPP). Major applications of the MPP are in the area of image processing (where the operands are often very small integers) from very high spatial resolution passive image sensors, signal processing of radar data, and numerical modeling simulations of climate. The system can be programmed in assembly language or a high level language. Information on background, status, architecture, programming, hardware reliability, applications, and the MPP's development as a national resource for parallel algorithm research are presented in outline form.

Source record↗
NASA NTRS2022

A Dynamo Simulation Generating Saturn-Like Small Magnetic Dipole Tilts

Abstract Among planetary dynamos, the magnetic field of Saturn stands out in its exceptional level of axisymmetry. One of its peculiar features is that the magnetic dipole mode is tilted with respect to the planetary rotation axis by only ≈0.007° or less. Numerical dynamo simulations performed in this context have had great difficulty in producing such small dipole tilt angles without introducing ad hoc ingredients such as a latitudinally varying heat flux pattern in the outer layers or stably stratified layers (SSLs). Here we present a numerical dynamo simulation that generates a highly axisymmetric dynamo with a dipole tilt of about ≈0.0008° on average. The model consists of a deep dynamo layer and an overlying low-conductivity layer but without any SSLs. We highlight a novel mechanism where strong differential rotation generated in the atmospheric layer penetrates into the dynamo region, helping to maintain a very small magnetic dipole tilt. Plain Language Summary Saturn's dipole-dominant magnetic field exhibits a very peculiar feature: the dipole component of the planetary magnetic field is tilted by less than ≈0.007° with respect to the planetary spin axis. Numerical simulations performed in this context suggest that if a spatial heat-flux variation is imposed, along with a stably stratified region, on top of an active dynamo layer, then small dipole tilt values can be realized. Here we present a model where extremely small dipole tilt values can be achieved without these ad hoc ingredients. Our simulations demonstrate that dynamo theory allows extremely small dipole tilt values in a relatively simple model configuration.

Rakesh Kumar Yadav↗
DOE OSTI2026

Data Summarization and Inference at Scale

This is the final report for the DOE ASCR grant SC-0022260, Data Summarization and Inference at Scale, PI: Alex Pothen, Purdue University. The goal of the project was to solve data-intensive and compute-intensive problems in the physical sciences, engineering, information science, data science, etc. by designing and implementing new algorithms that could work with a subset of the data. The four subgoals were: (a) The solution of problems where the data is too large to be stored in the memory of a computer. In this streaming model of computation, the data arrives as a stream of elements to the computer, each element is processed as it arrives, and a decision is made to discard the data or to store it; only a small subset of the data proportional to the size of the output solution is stored, and when all the data has been streamed, a solution to the problem is computed from the stored subset. (b) The use of machine learning methods to compute solutions to data-intensive problems. The use of GPUs is critical to obtain high performance on machine learning tasks, but their memory sizes are smaller relative to that of CPUs. For large-scale problems, the data is sampled many times, and small samples are used with repetition, for robustness, to compute solutions to inference tasks. This sampling reduces the memory required to solve the problem, but attention is needed to avoid slow convergence to the solutions, and reduced accuracy of inference. We propose submodular optimization, Large Language Models, and physics-informed neural networks to enable GPU computations here. (c) Modeling and visualization of high-dimensional data using interpretable features. Clinical proteomic data sets from immunology for the detection of cancer and other diseases are temporal and high-dimensional, and algorithms for visualizing these data sets using clinically interpretable features are lacking. We propose methods that compute distances based on the optimal transportation problem and graph edit distances to address this problem. We also propose the use of optimal transport-based distances, spatial statistics, and network structure to classify image data sets, We apply these algorithms to electron micrographs of the peripheral nervous system in the digestive tract. (d) The design of data-intensive algorithms on emerging architectures, specifically, noisy, intermediate-scale quantum (NISQ) devices. Quantum computers offer the possibility of exploring large solution spaces due to the principle of superposition, but current quantum computers are limited by few qubits, short coherence times due to noise, poor interconections among the qubits, etc. We propose the use of the divide and conquer paradigm to solve large-scale problems, wherein collections of small subproblems are solved on the quantum devices, and the solutions to the subproblems are integrated into a solution for the original problem on a classical computer.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1973

Small Interactive Image Processing System (SMIPS) system description

The Small Interactive Image Processing System (SMIPS) operates under control of the IBM-OS/MVT operating system and uses an IBM-2250 model 1 display unit as interactive graphic device. The input language in the form of character strings or attentions from keys and light pen is interpreted and causes processing of built-in image processing functions as well as execution of a variable number of application programs kept on a private disk file. A description of design considerations is given and characteristics, structure and logic flow of SMIPS are summarized. Data management and graphic programming techniques used for the interactive manipulation and display of digital pictures are also discussed.

Moik, J. G.↗
DOE OSTI2024

Prompt Phrase Ordering Using Large Language Models in HPC: Evaluating Prompt Sensitivity

Large language models (LLMs) have demonstrated effective performance in domain-specific tasks, often requiring a well-designed prompt to guide their responses. However, optimizing the right prompt is challenging due to prompt sensitivity—the phenomenon where small changes in the prompt can lead to significant variations in performance. In this study, we evaluate prompt performance by examining all permutations of independent phrases to investigate prompt sensitivity and robustness. We used two datasets: the GSM8k dataset, which assesses mathematical reasoning, and a custom template prompt for summarizing database metadata. Our goal was to evaluate the performance across all permutations of a sequence of prompt phrases. The study was conducted using the llama3-instruct- 7B model hosted on Ollama, with computations parallelized in a high-performance computing environment. By comparing the average index of phrases in the best and worst-performing prompts, we found that the order of independent phrases within a prompt significantly impacts LLM performance. Additionally, we used Hamming distance to assess changes between phrase orderings, concluding that prompt modifications can dramatically affect scores, often by almost random chance. These findings support existing research on prompt sensitivity. We discuss the challenges of prompt optimization, noting that altering phrases in a successful prompt does not always result in another successful prompt.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1978

Environment parameters and basic functions for floating-point computation

A language-independent proposal for environment parameters and basic functions for floating-point computation is presented. Basic functions are proposed to analyze, synthesize, and scale floating-point numbers. The model provides a small set of parameters and a small set of axioms along with sharp measures of roundoff error. The parameters and functions can be used to write portable and robust codes that deal intimately with the floating-point representation. Subject to underflow and overflow constraints, a number can be scaled by a power of the floating-point radix inexpensively and without loss of precision. A specific representation for FORTRAN is included.

Brown, W. S.↗