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.
As anyone even vaguely aware of current technology can tell you, machine learning (ML) and artificial intelligence (AI) have made exceptional breakthroughs in recent years. Generative artificial intelligence (GAI) emerged circa 2022 dominated by Large Language Models (LLMs) and generative tools for images emerged at about the same time.
Background: Suicide is a leading cause of death worldwide. Journalistic reporting guidelines were created to curb the impact of unsafe reporting; however, how suicide is framed in news reports may differ by important characteristics such as the circumstances and the decedent’s gender. Objective: This study aimed to examine the degree to which news media reports of suicides are framed using stigmatized or glorified language and differences in such framing by gender and circumstance of suicide. Methods: We analyzed 200 news articles regarding suicides and applied the validated Stigma of Suicide Scale to identify stigmatized and glorified language. We assessed linguistic similarity with 2 widely used metrics, cosine similarity and mutual information scores, using a machine learning–based large language model. Results: News reports of male suicides were framed more similarly to stigmatizing (P<.001) and glorifying (P=.005) language than reports of female suicides. Considering the circumstances of suicide, mutual information scores indicated that differences in the use of stigmatizing or glorifying language by gender were most pronounced for articles attributing legal (0.155), relationship (0.268), or mental health problems (0.251) as the cause.
Large language models (LLMs) are evolving from chatbots with limited tool-using capabilities to agentic AI systems that can perform deep research, assist in proposing hypotheses, help design experiments, automate data analysis, and draft scientific reports. However, there are currently two bottlenecks limiting LLMs' real-world impact on the broader scientific research community beyond academic demonstrations: lack of interoperability (repetitive manual tool-integration is required across scenarios) and the need for scalable coordination (unstructured communication and memory become brittle as the number of agents grows). In this Perspective, we argue that the next phase of agentic scientific discovery requires the development of an ecosystem of protocol-native agents and tools organized through hierarchies inspired by human society, beyond the current paradigm of a single monolithic “AI scientist”. We use Model Context Protocol (MCP) as a concrete example of an emerging interoperability layer for scientific tool and context exchange, and we propose three complementary pathways to increase the scaling capabilities of an MCP-native scientific ecosystem by addressing the composability issues: (1) MCP servers for high-value scientific tools maintained by domain experts, (2) automated transformation of existing code repositories into MCP services, and (3) autonomous invention and evolution of new agents and workflows. Finally, we provide a practical roadmap for scaling AI-driven scientific discovery by expanding tool supply and coordination in MCP-native scientific ecosystems.
As large language models (LLMs) continue to evolve, their capacity to replace humans as their surrogates is also improving. As increasing numbers of intelligent tutoring systems (ITSs) are embracing the integration of LLMs for digital tutoring, questions are arising as to how effective they are and if their hallucinatory behaviors diminish their perceived advantages. One critical question that is seldom asked if the availability, plurality, and relative weaknesses in the reasoning process of LLMs are contributing to the much discussed digital divide and equity and fairness in online learning. In this paper, we present an experiment with database design theory assignments and demonstrate that while their capacity to reason logically is improving, LLMs are still prone to serious errors. We demonstrate that in online learning and in the absence of a human instructor, LLMs could introduce inequity in the form of “wrongful” tutoring that could be devastatingly harmful for learners, which we call ignorant bias, in increasingly popular digital learning. We also show that significant challenges remain for STEM subjects, especially for subjects for which sound and free online tutoring systems exist. Based on the set of use cases, we formulate a possible direction for an effective ITS for online database learning classes of the future.
The integration of Large Language Models (LLMs) into scientific research promises to accelerate discovery, yet a significant gap remains between the dynamic reasoning of Artificial Intelligence (AI) agents and the static, deterministic nature of canonical scientific workflows. This paper introduces ADEPT (Agentic Discovery and Exploration Platform for Tools), a reference architecture and pedagogical framework explicitly designed to bridge this gap. ADEPT's primary mission is to provide a transparent, "glass-box" environment where researchers and engineers can learn to effectively wrap established scientific software (e.g., BLAST, Nextflow pipelines) and compose it into reliable, agent-driven workflows. We describe its modular, multi-server architecture, which leverages the Model Context Protocol (MCP) for tool serving, LangGraph for robust agentic orchestration, and a secure nsjail-based sandbox for safe code execution. By prioritizing architectural clarity, safety, and modularity, ADEPT serves as an extensible blueprint for building trustworthy AI-augmented systems and fosters the collaborative development necessary to responsibly employ agentic AI for science. We provide practical examples of how to adapt and extend this framework, highlighting its utility in workforce development and AI-readiness capabilities across research and development projects.
Plant bioengineering has generated tens of thousands of genotype-to-phenotype relationships, but this knowledge remains fragmented across narrative literature and difficult to use computationally. Inconsistent descriptions of DNA constructs, host species, and traits, including variable species names, omitted regulatory elements, and inconsistent gene symbols, impede data reuse, comparative analysis, and design-build-test-learn cycles. Here, we present the Plant Bioengineering Atlas, a literature-mined, ontology-grounded knowledge base assembled using an artificial intelligence (AI)-aided extraction pipeline. A large language model parsed open-access primary research articles to generate structured, provenance-anchored records of engineered genes, modification types, promoter-gene-terminator constructs, host species, target traits, and reported phenotypes, with every record traceable to its source. The current release contains 14,358 curated records encompassing 6,998 distinct genes across 436 plant species from 6,452 papers published between 2000 and 2026. Corpus analysis reveals that experiments are concentrated in a small group of model and crop species, disease and pathogen resistance is the most frequently engineered trait class, and constitutive regulatory parts (particularly the CaMV 35S promoter and NOS terminator) remain pervasive. Two in five records omit one or both flanking regulatory elements (i.e., promoter and terminator), while only 23.4% describe cassettes in which both elements resolve to named part classes, exposing a systematic reproducibility gap. We organize these data into a knowledge graph linking genes, constructs, species, and traits; provide access through an interactive web portal; and propose an AI-compatible documentation standard for AI-ready reporting. The Plant Bioengineering Atlas provides a foundation for data-driven hypothesis generation and AI-aided plant biodesign.
Motifs are distinctive, recurring, widely used idiom-like words or phrases, often originating from folklore, whose meaning is anchored in a narrative and have a significance as communicative devices across a wide range of media, including news, literature, and propaganda. Many motifs concisely imply a large constellation of culturally relevant information, and their broad usage suggests their cognitive importance as touchstones of cultural knowledge. As such, their detection is a step towards culturally aware natural language processing. We present GOLEM (GOld standard for Learning and Evaluation of Motifs) a dataset of English news articles, opinion pieces, and broadcast transcripts annotated for motific information. The dataset identifies 25,737 motif candidates across 34 motif types drawn from three cultural or national groups: Jewish, Irish, and Puerto Rican. The dataset contains 2,024,141 words split into 25,737 text snippets drawn from 8,073 articles. Each motif candidate is labeled according to a scheme which identifies the type of usage (motific, referential, eponymic, or unrelated), resulting in 1,743 actual motific instances in the data. Annotation was performed by individuals identifying as members of each group and achieved a Fleiss’ kappa (?) of > 0.55. In addition to the data, we demonstrate that classification of the candidate type is a challenging task for Large Language Models (LLMs) using a few-shot approach; recent models such as T5, FLAN-T5, GPT-2, and Llama 2 (7B) achieved a performance of 41% accuracy at best, where the majority class accuracy is 41% and the average chance accuracy is 27%. These data will support development of new models and approaches for detecting (and reasoning about) motific information in text.
As particle accelerators grow in complexity, traditional control methods face increasing challenges in achieving optimal performance. This paper envisions a paradigm shift: a decentralized multi-agent framework for accelerator control, powered by Large Language Models (LLMs) and distributed among autonomous agents. We present a proposition of a self-improving decentralized system where intelligent agents handle high-level tasks and communication and each agent is specialized control individual accelerator components. This approach raises some questions: What are the future applications of AI in particle accelerators? How can we implement an autonomous complex system such as a particle accelerator where agents gradually improve through experience and human feedback? What are the implications of integrating a human-in-the-loop component for labeling operational data and providing expert guidance? We show two examples, where we demonstrate viability of such architecture.
Large Language Models (LLMs), such as ChatGPT, tend to “hallucinate”, meaning they confidently generate false information. Retrieval Augmented Generation (RAG) attempts to diminish hallucination by providing context to the LLM from data stores (indexes) containing relevant information. The LLM uses this context to formulate its response. RAG systems can still suffer from hallucination because of bad embeddings or ineffective routing. For example, a router will often return context from an irrelevant index, resulting in a hallucinated answer. In this study, we aim to minimize the frequency of routing hallucinations by optimizing Index Summary Routing.
Vulnerability and patch management is an integral part of a robust cybersecurity program, yet it grows increasingly complex due to the sheer amount of data that must be analyzed. Particularly in Operational Technology (OT) environments, analysis must be done manually because of the lack of automated solutions. Additionally, there are many steps in this process, from the initial discovery of the vulnerability to the implementation of its remediation, and each step in the process requires different data in order to be performed effectively. In this work, we provide approaches and strategies to assist operators in industrial or OT environments throughout the vulnerability management cycle. Security advisories provide key information about mitigation strategies, or actions that can be taken when a patch is unavailable or cannot be installed. Details of these strategies are not shared in public vulnerability databases and must be found manually. We approach this problem by designing a solution to automatically identify that information within vendor security advisories and retrieve it for operator use. We start with an approach that requires domain-specific knowledge of certain frequently-seen reference websites. Next, an approach that can work on an arbitrary website but relies on certain keywords. Finally, an approach that uses Natural Language Processing (NLP) methods and does not require specific knowledge or keywords. Each of these approaches is more general than its predecessor; we demonstrate high accuracy for all approaches Advisories also often contain details of affected products in non-standard or natural language formats. While this information can be easily understood when read by an operator, the non-standard format acts as a barrier to effective automation. We provide an approach for the first step in this process: identifying vendors in security advisories and mapping them to a standard framework for representing digital assets and software products. We evaluate five established string similarity algorithms, plus one of our own design that combines string similarity and information theory, on the task of mapping vendors to their corresponding entries in the Common Platform Enumeration (CPE) repository. Our results show that our proposed metric outperforms all others. Due to the constraints on time, finances, and personnel for organizations, Large Language Models (LLMs) may seem like attractive opportunities for security operators to speed up information gathering; however, it is still not clear whether LLMs can handle vulnerability management tasks well. To answer this question, we perform an empirical study of LLMs’ ability to provide consistent, accurate information about vulnerabilities in order to guide organizations in their adoption of LLMs. We observe poor performance for all models tested, suggesting that these models are not well-suited to the consistent retrieval of accurate vulnerability information. Finally, once vulnerabilities have been identified and any additional information has been obtained, operators must decide which remediation actions to implement based on their available resources. This already-complex problem becomes even more so when we consider that a vulnerability may have multiple avenues for remediation. We formulate this scenario as two knapsack problems and provide solutions, which we then compare against several existing strategies for vulnerability prioritization seen in real operational environments.
Many workflows in high-energy-physics (HEP) stand to benefit from recent advances in transformer-based large language models (LLMs). While early applications of LLMs focused on text generation and code completion, modern LLMs now support orchestrated agency: the coordinated execution of complex, multi-step tasks through tool use, structured context, and iterative reasoning. We introduce the HEP Toolkit for Agentic Planning, Orchestration, and Deployment (HEPTAPOD), an orchestration framework designed to bring this emerging paradigm to HEP pipelines. The framework enables LLMs to interface with domain-specific tools, construct and manage simulation workflows, and assist in common utility and data analysis tasks through schema-validated operations and run-card-driven configuration. To demonstrate these capabilities, we consider a representative Beyond the Standard Model (BSM) Monte Carlo validation pipeline that spans model generation, event simulation, and downstream analysis within a unified, reproducible workflow. HEPTAPOD provides a structured and auditable layer between human researchers, LLMs, and computational infrastructure, establishing a foundation for transparent, human-in-the-loop systems.
To reduce the cost and time needed for regulatory compliance, nuclear power plants (NPPs) can utilize artificial intelligence (AI) to assist in interpreting complex and voluminous documents that typically span thousands of pages. Usually, the process of interpreting a plant’s technical specifications (TSs) and associated documents is labor intensive. This study aims to understand what processes state-of-the-art large language models (LLMs) can automate and to identify the pitfalls associated with using LLMs to reduce human labor costs and time. This research uses a recent AI technology called retrieval augmented generation (RAG), which retrieves pages of information from TSs and associated documents to assist with NPP power uprates (cleared to produce more power). LLMs are integral to RAG because they create human-like responses based on the retrieved information, aiding in the interpretation and application processes. A baseline case demonstrates how LLMs can operate successfully for a power uprate application. Then five use cases show five types of potential failures: (1) RAG retrieving the incorrect information, (2) RAG misinterpreting the retrieved information, (3) RAG relying on knowledge not contained in the retrieved information, (4) RAG hallucinating, and (5) RAG refusing to answer. The results of the five use cases suggest that automating the human interpretation of TSs and associated documents with AI should be approached with caution. A subject-matter expert reviewed the AI outputs from the five use cases and concluded that an LLM can produce technical information that is needed to produce power uprate applications in certain instances.
In cross-silo federated learning (FL), sensitive text datasets remain confined to local organizations due to privacy regulations, making repeated training for each downstream task both communication-intensive and privacy-demanding. A promising alternative is to generate differentially private (DP) synthetic datasets that approximate the global distribution and can be reused across tasks. However, pretrained large language models (LLMs) often fail under domain shift, and federated finetuning is hindered by computational heterogeneity: only resource-rich clients can update the model, while weaker clients are excluded, amplifying data skew and the adverse effects of DP noise. We propose a flexible participation framework that adapts to client capacities. Strong clients perform DP federated finetuning, while weak clients contribute through a lightweight DP voting mechanism that refines synthetic text. To ensure the synthetic data mirrors the global dataset, we apply control codes (e.g., labels, topics, metadata) that represent each client’s data proportions and constrain voting to semantically coherent subsets. This two-phase approach requires only a single round of communication for weak clients and integrates contributions from all participants. Experiments show that our framework improves distribution alignment and downstream robustness under DP and heterogeneity.
Operational safety in mission-critical environments requires AI systems that are accurate, interpretable, and resistant to hallucination. We present an agentic Retrieval-Augmented Generation (RAG) framework, REFSafe, for grounded hazard analysis and automated safety report generation. The system integrates Large Language Models (LLMs) with structured operational data, historical incident repositories, policy documents, and external authoritative sources. Through iterative agentic reasoning, the framework retrieves, verifies, and synthesizes evidence prior to generation, enforcing citation-backed outputs with explicit source attribution (documents, links, and prior events) to ensure traceability and trust. To mitigate hallucinations and unsupported claims, all risk assessments and forecasts are constrained to retrieved evidence, with confidence signals derived from retrieval relevance and source consistency. A transparent pipeline enables subject matter experts (SMEs) to validate predictions, and provide structured feedback, forming a continuous performance calibration loop. Preliminary deployment demonstrates improved reliability in hazard detection and safety/vulnerability report generation. This work advances trustworthy, evidence-grounded AI for predictive safety intelligence in mission-critical operations.
Retrieval-augmented generation (RAG) has emerged as a promising paradigm for improving factual accuracy in large language models (LLMs). We introduce a benchmark designed to evaluate RAG pipelines as a whole, evaluating a pipelines ability to ingest several modalities of information. We present (1) a curated dataset of 93 questions designed to evaluate a pipeline's ability to ingest textual data, tables, images, multimodal data, and cross-document multimodal data; (2) a phrase-level recall metric for correctness; (3) a nearest-neighbor embedding classifier in an attempt to classify pipeline hallucinations; (4) a comparative evaluation of 2 pipelines built with open-source retrieval mechanisms and 4 closed-source foundational models; and (5) a third-party human evaluation of the alignment of our correctness and hallucination metrics. We find that closed-source pipelines significantly outperform open-source pipelines in both the correctness and halucination metrics, with a wider performance gap in questions relying on multimodal and cross-document information. We also find after a human evaluation of our correctness and hallucination metric compared with our questions and pipeline responses, average agreement was 4.62 for correctness 4.53 for hallucination detection on a 1-5 Likert scale with 5 being strongly agree with our determination.
Operational safety in high-stakes domains such as industrial process control, autonomous, and safety-critical systems demand reliable hazard identification. While large language models (LLMs) have shown promise in automating safety analysis tasks, single-turn, monolithic inference is brittle: it lacks the self-correction, deliberation, and contextual refinement that safety engineers apply iteratively. In this paper, we introduce HAZDIAL, a framework that investigates whether structured agentic dialogue (multi-agent, multi-turn interactions) improves the quality of NLP-based hazard identification over single-pass baselines. We systematically compare two dialogue modalities: adversarial debate and constructive discussion, and propose an genetic algorithm-based agentic interaction optimization. We evaluate all configurations against a curated golden dataset using standard classification metrics (accuracy, precision, recall, F1) and a novel dialogue metrics. This work advances the intersection of dialogue systems, multi-agent reasoning, and AI safety, providing empirical evidence for dialogue-driven hazard analysis.
Agentic systems now address tasks across theoretical, phenomenological, and experimental high energy physics (HEP), but their scientific capabilities remain difficult to reuse across different large language models, providers, and harnesses. We argue that stable parts of these workflows should be promoted into versioned scientific operations and exposed through common protocols. Existing general-purpose harnesses can then be specialized for HEP through task-specific sets of tools and skills, while community-maintained registries would make these capabilities discoverable and citable. We identify mismatches in conventions, assumptions, and domains of validity among independently developed operations as a potential obstacle to their composition, and discuss machine-readable scientific contracts as one possible solution. These design principles and evaluation guidelines provide a near-term path toward a portable and community-maintained agentic harness for HEP.