Arbitrary Autoencoder Injection for Interpretability Experimentation
Goes over a simple software library (Python) for utilizing sparse autoencoders for more models than just language models.
Campos, Marco Vinicio↗
SEARCH · Engineering Papers
Results for “language models”
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.
Generative AI in Supply Chain Management: Applications, Challenges, and Future Directions
Supply chain management (SCM) is undergoing rapid transformation due to increasing global complexity, demand volatility, and operational disruptions. Generative Artificial Intelligence (GenAI) has emerged as a powerful paradigm capable of synthesizing data, simulating operational scenarios, and enabling adaptive decision-making across supply chain networks. This paper presents a survey of GenAI’s role in SCM, focusing on its applications in predictive analytics, autonomous logistics, and fraud detection. Unlike traditional AI systems that rely primarily on predictive analytics, GenAI models, including large language models, generative adversarial networks, and diffusion-based architectures, enable the creation of synthetic supply chain scenarios and autonomous optimization strategies. This survey provides (1) a taxonomy of GenAI techniques for supply chain applications, (2) a comparative analysis of generative AI approaches with traditional machine learning, reinforcement learning, and blockchain-based methods, and (3) a discussion of key challenges such as data privacy, interpretability, and integration with legacy enterprise systems. Furthermore, we outline open research problems and propose directions for future research toward autonomous, resilient, and sustainable AI-driven supply chains.
15 - GEOTHERMAL ENERGY↗
Towards philosophical reasoning with agentic LLMs: Socratic method for scientific assistance
As large language models (LLMs) become central tools in science, improving their reasoning capabilities is critical for meaningful and trustworthy applications. We introduce a Socratic agent for scientific reasoning, implemented through a structured system prompt that guides LLMs via classical principles of inquiry. Unlike typical prompt engineering or retrieval-based methods, our approach leverages definition, analogy, hypothesis elimination, and other Socratic techniques to generate more coherent, critical, and domain-aware responses. We evaluate the agent across diverse scientific domains and benchmark it on the abstraction and reasoning corpus challenge dataset, achieving 97.15% under a fixed prompting protocol and without fine-tuning or external tools. Expert evaluation shows improved reasoning depth, clarity, and adaptability over conventional LLM outputs, suggesting that structured prompting rooted in philosophical reasoning can improve the scientific utility of language models.
LLM reasoning↗
MSD CoP Webinar: "Generative agents: A new frontier for representing human actors and their behavior in MSD models"
Context: This webinar was hosted by the MultiSector Dynamics Community of Practice (MSD CoP; https://multisectordynamics.org). Talk #1: Behavioral Generative Agents for Energy Operations Presenter: Dr. Cong Chen (Thayer School of Engineering, Dartmouth College) Abstract: Accurately modeling consumer behavior in energy operations remains challenging due to inherent uncertainties, behavioral complexities, and limited empirical data. This talk introduces a novel approach leveraging generative agents--artificial agents powered by large language models--to realistically simulate customer decision-making in dynamic energy operations. Talk #2: Simulating multiple human perspectives in socio-ecological systems using large language models Presenter: Dr. Yongchao Zeng (Institute of Meteorology and Climate Research, Atmospheric Environmental Research (IMK-IFU) of the Karlsruhe Institute of Technology in Germany) Abstract: Understanding socio-ecological systems requires insights from diverse stakeholder perspectives. This talk describes a novel simulation system called HoPeS (Human-oriented Perspective Shifting). HoPeS enables model users to not only explore simulated socio-ecological systems (SESs) from a third-person observer's perspective but also take any of the simulated stakeholder roles, like playing an RPG game. By shifting multiple perspectives, model users can reflect and integrate the situated knowledge learned through the participatory simulation, approximating a more holistic and less biased understanding of SESs. Moderators: Jim Yoon (MSD CoP Human Systems Modeling Working Group Co-Chair); Stefano Galelli (MSD CoP Using AI to Enhance MSD Research Working Group Co-Chair); Patrick M. Reed (MSD CoP Facilitation Team) This webinar was held on: November 13th, 2025 from 12-1 PM EST.
Artificial Intelligence↗
Improving Synonym Recommendation Using Sentence Context
Traditional synonym recommendations often include ill-suited suggestions for writer's specific contexts. We propose a simple approach for contextual synonym recommendation by combining existing human-curated thesauri, e.g. WordNet, with pre-trained language models. We evaluate our technique by curating a set of word-sentence pairs balanced across corpora and parts of speech, then annotating each word-sentence pair with the contextually appropriate set of synonyms. We found that basic language model approaches have higher precision. Approaches leveraging sentence context have higher recall. Overall, the latter contextual approach had the highest F-score.
Glenski, Maria F.↗
How Beneficial Is Pretraining on a Narrow Domain-Specific Corpus for Information Extraction about Photocatalytic Water Splitting?
Language models trained on domain-specific corpora have been employed to increase the performance in specialized tasks. However, little previous work has been reported on how specific a “domain-specific” corpus should be. Here, we test a number of language models trained on varyingly specific corpora by employing them in the task of extracting information from photocatalytic water splitting. We find that more specific corpora can benefit performance on downstream tasks. Furthermore, PhotocatalysisBERT, a pretrained model from scratch on scientific papers on photocatalytic water splitting, demonstrates improved performance over previous work in associating the correct photocatalyst with the correct photocatalytic activity during information extraction, achieving a precision of 60.8(+11.5)% and a recall of 37.2(+4.5)%.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Cybersecurity-Tools/V2W-BERT
Software Vulnerabilities to Weakness Mapping. V2W-BERT optimizes the learnable parameters θ of Fθ in two steps. In the first step, the pre-trained BERT language model is further fine-tuned with CVE/CWE descriptions specific to cyber security. In the second step, the trained BERT model is employed in a Siamese network architecture to establish links between CVEs and CWEs. The architecture takes a specific CVE-CWE pair as input and predicts whether the CVE belongs to the CWE or not, with a confidence value. V2W-BERT includes a Mask Language Model (LM) based Reconstruction Decoder to ensure that the descriptions’ contexts are not changed too much during the training process.
Halappanavar,, Mahantesh↗
LLaMP v0.1.0
Reducing hallucination of Large Language Models (LLMs) is imperative for use in the sciences, where reliability and reproducibility are crucial. However, LLMs inherently lack long-term memory, making it a nontrivial, ad hoc, and inevitably biased task to fine-tune them on domain-specific literature and data. LLaMP is a multimodal retrieval-augmented generation (RAG) framework of hierarchical reasoning and acting (ReAct) agents that can dynamically and recursively interact with Materials Project to ground large language models on high-fidelity materials informatics.
Riebesell, Janosh [Lawrence Berkeley National Labo↗
Utilizing Machine Learning to Improve Neutralization Potency of an HIV-1 Antibody Targeting the gp41 N-Heptad Repeat
The N-heptad repeat (NHR) of the HIV-1 gp41 prehairpin intermediate (PHI) is an attractive potential vaccine target with high sequence conservation across diverse strains. However, despite the potency of NHR-targeting peptides and clinical efficacy of the NHR-targeting entry inhibitor enfuvirtide, no potently neutralizing NHR-directed monoclonal antibodies (mAbs) nor antisera have been identified or elicited to date. The lack of potent NHR-binding mAbs both dampens enthusiasm for vaccine development efforts at this target and presents a barrier to performing passive immunization experiments with NHR-targeting antibodies. To address this challenge, we previously developed an improved variant of the NHR-directed mAb D5, called D5_AR, which is capable of neutralizing diverse tier-2 viruses. Building on that work, here we present the 2.7Å-crystal structure of D5_AR bound to NHR mimetic peptide IQN17. We then utilize protein language models and supervised machine learning to generate small (n < 100) libraries of D5_AR variants that are subsequently screened for improved neutralization potency. We identify a variant with 5-fold improved neutralization potency, D5_FI, which is the most potent NHR-directed monoclonal antibody characterized to date and exhibits broad neutralization of tier-2 and −3 pseudoviruses as well as replicating R5 and X4 challenge strains. Additionally, our work highlights the ability of protein language models to efficiently identify improved mAb variants from relatively small libraries.
Biopolymers↗
U.S. Solar Siting Regulation and Zoning Ordinances (2025)
A machine readable collection of documented solar siting ordinances at the state and local (e.g., county, township) level throughout the United States. The data were compiled using the Infrastructure Continuous Ordinance Mapping for Planning and Siting Systems (INFRA-COMPASS) tool, which leverages Large Language Models (LLMs) to automate the collection of local codes and ordinances applicable to energy infrastructure. URLs for the ordinance source documents are included in the Solar Ordinances spreadsheet. The GeoPackage file included below contains the jurisdiction shapes for each ordinance. Note that the GeoPackage file is formatted for ingestion by NLR's reVX setbacks tool and therefore does not contain any of the state-level regulations. NOTE: This data was collected with the help of generative AI. The Large Language Models used for this effort make mistakes. Always validate the data for critical use cases. This data is an update to a previously developed database of wind ordinances found in OEDI Submission 5734: see the "U.S. Solar Siting Regulation and Zoning Ordinances 2022" link below. INFRA-COMPASS version used for collection: v0.11.3 LLMs used for collection: GPT-4.1, GPT-4.1 mini, GPT-4.1 nano
14 SOLAR ENERGY↗
U.S. Wind Siting Regulation and Zoning Ordinances (2025)
A machine readable collection of documented wind siting ordinances at the state and local (e.g., county, township) level throughout the United States. The data were compiled using the Infrastructure Continuous Ordinance Mapping for Planning and Siting Systems (INFRA-COMPASS) tool, which leverages Large Language Models (LLMs) to automate the collection of local codes and ordinances applicable to energy infrastructure. URLs for the ordinance source documents are included in the Wind Ordinances spreadsheet. The GeoPackage file included below contains the jurisdiction shapes for each ordinance. Note that the GeoPackage file is formatted for ingestion by NREL's reVX setbacks tool and therefore does not contain any of the state-level regulations. NOTE: This data was collected with the help of generative AI. The Large Language Models used for this effort make mistakes. Always validate the data for critical use cases. This data is an update to a previously developed database of wind ordinances found in OEDI Submission 5733: see the "U.S. Wind Siting Regulation and Zoning Ordinances 2022" link below. INFRA-COMPASS version used for collection: v0.8.2 LLMs used for collection: GPT-4.1, GPT-4.1 mini, GPT-4.1 nano, GPT-4o mini
17 WIND ENERGY↗
Agentic AI vs ML-Based Autotuning: A Comparative Study for Loop Reordering Optimization
High Performance Computing (HPC) applications rely heavily on code optimizations to achieve good performance on modern CPU and GPU architectures. Traditional Machine Learning auto-tuning approaches have demonstrated success in exploring high-dimensional spaces, but they often require expensive compile-run evaluations and lack adaptability for large HPC applications. The recent advances in Large Language Models (LLMs) and Agentic AI systems raise intriguing questions about the potential of these approaches to address specific optimization methodologies. This work aims to answer an essential question for the HPC community: “How Agentic AI Systems Compare to Traditional ML Autotuning Techniques?” To address this question, we present a comparative analysis between a traditional ML-based optimization approach and an Agentic AI system, evaluating their respective capabilities and limitations for loop-level optimization. In addition, we introduced a new Agentic AI system named LoopGen-AI using three different Large Language Models: GPT-4.1, Claude 4.0, and Gemini 2.5. A key finding is that LoopGen-AI achieves competitive per-formance with only a few program runs, the reasoning logs from the agents revealed that their decisions rely heavily on the combination of semantic understanding of the target kernel with dynamic feedback from the environment, highlighting a promising new dimension in performance tuning. In contrast, ML-based autotuners focus on statistical exploration, and require orders of magnitude more runs to reach peak performance. Additionally, our analysis shows that prompt engineering, particularly using Persona + Context Manager patterns, significantly impacts the effectiveness of Agentic AI. Our results indicate that while Agentic AI systems are not yet a complete replacement for ML-based autotuners, it can effectively complement traditional methods.
Rosas, Miguel Romero↗
From natural language to control signals: a conceptual framework for semantic channel finding in complex experimental infrastructure
Modern experimental platforms such as particle accelerators, fusion devices, telescopes, and industrial process control systems expose tens to hundreds of thousands of control and diagnostic channels, accumulated over decades of hardware evolution. Operators and AI systems alike depend on informal expert knowledge, inconsistent naming conventions, and scattered documentation to locate the signals required for monitoring, troubleshooting, and automated control, creating a persistent bottleneck for reliability, scalability, and emerging language-model-driven interfaces. We formalize semantic channel finding, the task of mapping natural-language intent to concrete control-system signals, as a general problem in complex experimental infrastructure, and introduce a four-paradigm conceptual framework to guide architecture selection based on facility-specific data regimes. The paradigms span (i) direct in-context lookup over small, curated channel dictionaries, (ii) constrained hierarchical navigation through structured trees, (iii) interactive agent exploration using iterative reasoning and tool-based database queries, and (iv) ontology-grounded semantic search that decouples channel meaning from facility-specific naming conventions. We demonstrate the practical feasibility of each paradigm through proof-of-concept implementations at four operational facilities spanning two orders of magnitude in scale: from compact free-electron lasers to large synchrotron light sources, operating under diverse control-system architectures ranging from clean hierarchical naming schemes to legacy environments with decades of heterogeneous conventions. Where evaluated against expert-curated operational queries, these instantiations achieve 90%–97% accuracy, validating the framework’s applicability across real-world deployment scenarios. To accelerate adoption across the broader scientific and industrial control-system community, we release open-source, plug-and-play implementations of all three interactive paradigms-direct lookup, hierarchical navigation, and middle-layer exploration-within the Osprey framework, together with tools for channel database generation, interactive testing, and minimal-configuration deployment. This work establishes semantic channel finding as a foundational capability for human-centric and agentic AI interfaces at large-scale facilities, providing both a systematic framework for architecture design and practical resources to enable adoption without building custom infrastructure from scratch.
channel finding↗
VWC-BERT: Scaling Vulnerability–Weakness–Exploit Mapping on Modern AI Accelerators
Defending cybersystems needs accurate mapping of software and hardware vulnerabilities to generalized descriptions of weaknesses, and weaknesses to exploits. These mappings enable cyber defenders to build plans for effective defense and assessment of potential risks to a cybersystem. With close to 170k vulnerabilities, manual mapping is not a feasible option. However, automated mapping is challenging due to limited training data, computational intractability, and limitations in computational natural language processing. Tools based on breakthroughs in Transformer-based language models have been demonstrated to classify vulnerabilities with high accuracy. We make three key contributions in this paper: (1) We present a new framework, \VWCBERT, that augments the Transformer-based hierarchical multi-class classification framework of Das et al. (\textsc{V2W-BERT}) with the ability to map weaknesses to exploits. (2) We implement \VWCBERT~ on modern AI accelerator platforms using two data parallel techniques for the pre-training phase and demonstrate nearly linear speedups across NVIDIA and Graphcore accelerator platforms. We observe nearly linear speedups for up to 16 V100 and 8 A100 GPUs, and about 3.4$\times$ speedup for A100 relative to V100 GPUs. We also observe excellent speedups on Graphcore, with $5.7\times$ speedup on 128 IPUs relative to 16 IPUs. Enabled by scaling, we also demonstrate higher accuracy using a larger language model, RoBERTa-Large. We show up to 87\% accuracy for strict and up to 98\% accuracy for relaxed classification. (3) We develop a novel parallel link manager for the link prediction phase and demonstrate up to 21$\times$ speedup with 16 V100 GPUs relative to one V100 GPU, and thus reducing the runtime from 2.5 hours to 10 minutes. We believe that generalizability and scalability of \VWCBERT~ will benefit both the theoretical development and practical deployment of novel cyberdefense solutions and vulnerability classification.
Das, Siddhartha Shankar↗
RoadBench: A Vision-Language Foundation Model and Benchmark for Road Damage Understanding
Explore the source record for details and available documents.
Xiao, Xi [ORNL] (ORCID:0009000009316982)↗
FM4NPP: A Scaling Foundation Model for Nuclear and Particle Physics
Large language models have revolutionized artificial intelligence by enabling large, generalizable models trained through self-supervision. This paradigm has inspired the development of scientific foundation models (FMs). However, applying this capability to experimental particle physics is challenging due to the sparse, spatially distributed nature of detector data, which differs dramatically from natural language. This work addresses if an FM for particle physics can scale and generalize across diverse tasks. We introduce a new dataset with more than 11 million particle collision events and a suite of downstream tasks and labeled data for evaluation. We propose a novel self-supervised training method for detector data and demonstrate its neural scalability with models that feature up to 188 million parameters. With frozen weights and task-specific adapters, this FM consistently outperforms baseline models across all downstream tasks. The performance also exhibits robust data-efficient adaptation. Further analysis reveals that the representations extracted by the FM are task-agnostic but can be specialized via a single linear mapping for different downstream tasks.
72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
SCIMON: Scientific Inspiration Machines Optimized for Novelty
We explore and enhance the ability of neu- ral language models to generate novel scien- tific directions grounded in literature. Work on literature-based hypothesis generation has traditionally focused on binary link prediction— severely limiting the expressivity of hypothe- ses. This line of work also does not focus on optimizing novelty. We take a dramatic depar- ture with a novel setting in which models use as input background contexts (e.g., problems, experimental settings, goals), and output natu- ral language ideas grounded in literature. We present SCIMON, a modeling framework that uses retrieval of “inspirations” from past scien- tific papers, and explicitly optimizes for novelty by iteratively comparing to prior papers and up- dating idea suggestions until sufficient novelty is achieved. Comprehensive evaluations reveal that GPT-4 tends to generate ideas with over- all low technical depth and novelty, while our methods partially mitigate this issue. Our work represents a first step toward evaluating and developing language models that generate new ideas derived from the scientific literature.
Ji, Heng↗
Privacy-Aware RAG-Enabled LLMs for Collaborative AI in Organizations
Recent advancements in Large Language Models (LLMs) based on Transformer architectures have significantly improved capabilities in natural language processing and generation. However, deploying LLMs for inter-organizational communication poses challenges, in ensuring privacy and facilitating effective collaboration. This paper introduces a novel decentralized inference meta-agent chatbot that leverages privacy-aware Retrieval-Augmented Generation (RAG)-enabled LLMs for collaborative AI communication across organizations. Built on Microsoft’s Autogen, the platform enables LLMs to autonomously refine responses, enhancing accuracy and relevance. It incorporates advanced hallucination mitigation techniques using Uptrain and a privacy-focused RAG framework that employs synthetic document generation to protect sensitive information. Comprehensive evaluations demonstrate the platform’s effectiveness in maintaining contextual relevance and stringent privacy standards, effectively addressing critical challenges in LLM-enhanced collaborative AI communication. This work represents a significant step toward secure and efficient inter-organizational collaboration using advanced generative AI technologies.
97 - MATHEMATICS AND COMPUTING↗