SEARCH · Engineering Papers
Results for “Large 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.
Interface To The SURE Program
Abstract Semi-Markov Specification Interface to SURE Tool (ASSIST) computer program is interface program enabling reliability engineers to design large semi-Markov mathematical models accurately. Language enables efficient description of large, complicated systems. Also offers, as part of bundled package with SURE and PAWS/STEM, two other reliable analysis programs developed by Systems Validation Methods group at Langley Research Center.
Foundation models for atomistic simulation of chemistry and materials
Conventional computational methods for modeling chemical and materials systems are limited by system size and timescale, forcing a trade-off between quantum-mechanical accuracy and the sampling needed for realistic observables. Large language and vision foundation models — pre-trained on massive datasets using transformer architectures — have revolutionized many fields. It is thus interesting to ask whether a foundation model — subject to suitable data, parameter scaling and training — could enable learned simulations of chemistry and materials. Here, in this study, we review the field of machine-learned interatomic potentials (MLIPs) and posit that scaling up large and diverse chemical and materials datasets and highly expressive architectures using advanced training strategies should result in models that are: more efficient, transferable, robust to out-of-distribution scenarios, and easier to fine-tune to a variety of downstream physical observables than models trained from scratch on small datasets corresponding to specific, targeted atomistic simulation tasks. We provide specific criteria for creating such large-scale MLIP foundation models, coordinated strategies for their development, evaluation and deployment, and highlight potential emergent capabilities that could transform predictive simulations in chemistry and materials science and accelerate discovery across multiple technological domains.
Automated generation of reliability models
The abstract semi-Markov specification interface to the SURE (Semi-Markov Range Evaluator) tool (ASSIST) program allows the user to describe the Markov model in a high-level language. Instead of listing the individual states of the model, the user specifies the rules governing the behavior of the system, and these are used to automatically generate the model. A small number of statements in the abstract language can describe a large, complex model. Becuase no assumptions are made about the system being modeled, ASSIST can be used to generate models describing the behavior of any type of system. The abstract model definition and the automatic model generation strategy are described. Analysis of an example fault-tolerant architecture, a triad of processor with cold spare processors, shows how the behavior of a system can be captured by a few general rules. The syntax of the ASSIST input language is then described and demonstrated by creating a model to describe the fault behavior of the example architecture. The flexibility of the abstract language is demonstrated by expanding the example to model multiple triads of processors sharing a pool of cold spare processors.
ASSIST user manual
Semi-Markov models can be used to analyze the reliability of virtually any fault-tolerant system. However, the process of delineating all the states and transitions in a complex system model can be devastatingly tedious and error prone. The Abstract Semi-Markov Specification Interface to the SURE Tool (ASSIST) computer program allows the user to describe the semi-Markov model in a high-level language. Instead of listing the individual model states, the user specifies the rules governing the behavior of the system, and these are used to generate the model automatically. A few statements in the abstract language can describe a very large, complex model. Because no assumptions are made about the system being modeled, ASSIST can be used to generate models describing the behavior of any system. The ASSIST program and its input language are described and illustrated by examples.
Flexible language constructs for large parallel programs
The goal of the research described is to develop flexible language constructs for writing large data parallel numerical programs for distributed memory (MIMD) multiprocessors. Previously, several models have been developed to support synchronization and communication. Models for global synchronization include SIMD (Single Instruction Multiple Data), SPMD (Single Program Multiple Data), and sequential programs annotated with data distribution statements. The two primary models for communication include implicit communication based on shared memory and explicit communication based on messages. None of these models by themselves seem sufficient to permit the natural and efficient expression of the variety of algorithms that occur in large scientific computations. An overview of a new language that combines many of these programming models in a clean manner is given. This is done in a modular fashion such that different models can be combined to support large programs. Within a module, the selection of a model depends on the algorithm and its efficiency requirements. An overview of the language and discussion of some of the critical implementation details is given.
TransPolymer: a Transformer-based language model for polymer property predictions
Accurate and efficient prediction of polymer properties is of great significance in polymer design. Conventionally, expensive and time-consuming experiments or simulations are required to evaluate polymer functions. Recently, Transformer models, equipped with self-attention mechanisms, have exhibited superior performance in natural language processing. However, such methods have not been investigated in polymer sciences. Herein, we report TransPolymer, a Transformer-based language model for polymer property prediction. Our proposed polymer tokenizer with chemical awareness enables learning representations from polymer sequences. Rigorous experiments on ten polymer property prediction benchmarks demonstrate the superior performance of TransPolymer. Moreover, we show that TransPolymer benefits from pretraining on large unlabeled dataset via Masked Language Modeling. Experimental results further manifest the important role of self-attention in modeling polymer sequences. We highlight this model as a promising computational tool for promoting rational polymer design and understanding structure-property relationships from a data science view.
Accelerating Transformer-based Deep Learning Models on FPGAs using Column Balanced Block Pruning
Although Transformer-based language representations achieve state-of-the-art accuracy on various natural language processing (NLP) tasks, the large model size has been challenging the resource constrained computing platforms. Weight pruning, as a popular and effective technique in reducing the number of weight parameters and accelerating the Transformer, has been investigated on GPUs. However, the Transformer acceleration using weight pruning on field-programmable gate array (FPGAs) remains unexplored. This paper investigates the column balanced block-wise pruning on Transformer and designs an FPGA acceleration engine to customize the balanced blockwise matrix multiplication. We implement the Transformer model with proper hardware scheduling, and the experiments show that the Transformer inference on FPGA achieves 10.35 ms latency with the batch size of 32, which is 10.96 × speed up comparing to CPU platform and 2.08 × speed up comparing to GPU platform.
SafeAeroBERT: Towards a Safety-Informed Aerospace-Specific Language Model
As aviation systems continue to operate with high traffic, large amounts of documents containing safety-relevant data continue to be generated via reporting systems such as the ASRS. Advanced natural language processing techniques, specifically pre-trained language models, have shown great success in domain-specific applications; however, the text in aviation safety reports is inundated with jargon and thus not fully utilized by general pre-trained models. In this research, we work towards developing a safety-informed aerospace-specific language model by pre-training a Bidirectional Encoder Representations from Transformer (BERT) model on reports from the Aviation Safety Reporting System and the National Transportation Safety Board. The resulting model, called SafeAeroBERT, is fine-tuned for the specific task of document classification, and can be further tuned for named-entity recognition, relation detection, information retrieval, and summarization. Results from the classification task are compared between SafeAeroBERT, the base BERT, and SciBERT models and show SafeAeroBERT outperforms the general BERT and SciBERT on classifying reports about human factors, aircraft, and procedure. SafeAeroBERT can be used on custom tasks, not limited to document classification, and is intended to aid an intelligent knowledge manager for safety report repositories.
SafeAeroBERT: Towards a Safety-Informed Aerospace-Specific Language Model
As aviation systems continue to operate with high traffic, large amounts of documents containing safety-relevant data continue to be generated via reporting systems such as the Aviation Safety Reporting System (ASRS). Advanced natural language processing techniques, specifically pre-trained language models, have shown great success in domain-specific applications; however, the text in aviation safety reports is inundated with jargon and thus not fully utilized by general pre-trained models. In this research, we work towards developing a safety-informed aerospace-specific language model by pre-training a Bidirectional Encoder Representations from Transformer (BERT) model on reports from the Aviation Safety Reporting System and the National Transportation Safety Board. The resulting model, called SafeAeroBERT, is fine-tuned for the specific task of document classification, and can be further tuned for named-entity recognition, relation detection, information retrieval, and summarization. Results from the classification task are compared between SafeAeroBERT, the base BERT, and SciBERT models and show SafeAeroBERT outperforms the general BERT and SciBERT on classifying reports about weather and procedure. SafeAeroBERT can be used on custom tasks, not limited to document classification, and is intended to aid an intelligent knowledge manager for safety report repositories.
A model for the distributed storage and processing of large arrays
A conceptual model for parallel computations on large arrays is developed. The model provides a set of language concepts appropriate for processing arrays which are generally too large to fit in the primary memories of a multiprocessor system. The semantic model is used to represent arrays on a concurrent architecture in such a way that the performance realities inherent in the distributed storage and processing can be adequately represented. An implementation of the large array concept as an Ada package is also described.
Automated model generation for reliability analysis programs
Semi-Markov models (a generalization of Markov models) can be used to calculate the reliability of virtually any fault-tolerant system. However, the process of delineating all of the states and transitions in the model of a complex system can be devastatingly tedious and error-prone. The ASSIST program allows the user to describe the semi-Markov model in a high-level language. Instead of specifying the individual states of the model, the user specifies the rules governing the behavior of the system, and these are used by ASSIST to automatically generate the model. A small number of statements in the abstract language can be used to describe a very large, complex model. Because no assumptions are made about the system being modeled, the ASSIST program can be used to generate models describing the behavior of any type of system. The ASSIST program and its input language are described and illustrated by examples.
A Modular, Portable Model of Image Fidelity
There is a persistent need for a trustworthy model of perceptual image fidelity, especially in applications such as image compression and display design. A fidelity model provides a measure of the visual discriminability of two images. Ahumada has previously shown that the existing fidelity models may be categorized according to their inclusion of various canonical properties, such as a contrast sensitivity function, spatial frequency channels, etc. This suggests that research would be aided by the availability of a modular model, in which these components could be easily inserted or removed. A further impediment to research in this area has been that most models are written in low-level languages and are consequently large, non-portable, and difficult to understand, modify, and maintain. We therefore believe research would also be aided by models written in high-level languages. To serve both of these purposes, and to honor our conference host for his lifetime dedication to the problem of image quality. Global brightness and its effect on perceptual image quality. We offer a modular model written in the high-level language Mathematica. We will demonstrate this model and show how it may be modified.
Concept-based Analysis of Neural Networks via Vision-Language Models
The analysis of vision-based deep neural networks (DNNs) is highly desirable but it is very challenging due to the difficulty of expressing formal specifications for vision tasks and the lack of efficient verification procedures. In this paper, we propose to leverage emerging multimodal, vision-language, foundation models (VLMs) as a lens through which we can reason about vision models. VLMs have been trained on a large body of images accompanied by their textual description, and are thus implicitly aware of high-level, human-understandable concepts describing the images. We describe a logical specification language Con spec designed to facilitate writing specifications in terms of these concepts. To define and formally check Con spec specifications, we build a map between the internal representations of a given vision model and a VLM, leading to an efficient verification procedure of natural-language properties for vision models. We demonstrate our techniques on a ResNet-based classifier trained on the RIVAL-10 dataset using CLIP as the multimodal model.
Domain-Specific Languages and Diagram Customization for a Concurrent Engineering Environment
A major open question for advocates of Model-Based Systems Engineering (MBSE) is the question of how system and subsystem engineers will work together. The Systems Modeling Language (SysML), like any language intended for a large audience, is in tension between the desires for simplicity and for expressiveness. In order to be more expressive, many specialized language elements may be introduced, which will unfortunately make a complete understanding of the language a more daunting task. While this may be acceptable for systems modelers, it will increase the challenge of including subsystem engineers in the modeling effort. One possible answer to this situation is the use of Domain-Specific Languages (DSL), which are fully supported by the Unified Modeling Language (UML). SysML is in fact a DSL for systems engineering. The expressive power of a DSL can be enhanced through the use of diagram customization. Various domains have already developed their own schematic vocabularies. Within the space engineering community, two excellent examples are the propulsion and telecommunication subsystems. A return to simple box-and-line diagrams (e.g., the SysML Internal Block Diagram) are in many ways a step backward. In order allow subsystem engineers to contribute directly to the model, it is necessary to make a system modeling tool at least approximate in accessibility to drawing tools like Microsoft PowerPoint and Visio. The challenge is made more extreme in a concurrent engineering environment, where designs must often be drafted in an hour or two. In the case of the Jet Propulsion Laboratory's Team X concurrent design team, a subsystem is specified using a combination of PowerPoint for drawing and Excel for calculation. A pilot has been undertaken in order to meld the drawing portion and the production of master equipment lists (MELs) via a SysML authoring tool, MagicDraw. Team X currently interacts with its customers in a process of sharing presentations. There are several inefficiencies that arise from this situation. The first is that a customer team must wait two weeks to a month (which is 2-4 times the duration of most Team X studies themselves) for a finalized, detailed design description. Another is that this information must be re-entered by hand into the set of engineering artifacts and design tools that the mission concept team uses after a study is complete. Further, there is no persistent connection to Team X or institutionally shared formulation design tools and data after a given study, again reducing the direct reuse of designs created in a Team X study. This paper presents the underpinnings of subsystem DSLs as they were developed for this pilot. This includes specialized semantics for different domains as well as the process by which major categories of objects were derived in support of defining the DSLs. The feedback given to us by the domain experts on usability, along with a pilot study with the partial inclusion of these tools is also discussed.
Adaptive language model training for molecular design
Abstract The vast size of chemical space necessitates computational approaches to automate and accelerate the design of molecular sequences to guide experimental efforts for drug discovery. Genetic algorithms provide a useful framework to incrementally generate molecules by applying mutations to known chemical structures. Recently, masked language models have been applied to automate the mutation process by leveraging large compound libraries to learn commonly occurring chemical sequences (i.e., using tokenization) and predict rearrangements (i.e., using mask prediction). Here, we consider how language models can be adapted to improve molecule generation for different optimization tasks. We use two different generation strategies for comparison, fixed and adaptive. The fixed strategy uses a pre-trained model to generate mutations; the adaptive strategy trains the language model on each new generation of molecules selected for target properties during optimization. Our results show that the adaptive strategy allows the language model to more closely fit the distribution of molecules in the population. Therefore, for enhanced fitness optimization, we suggest the use of the fixed strategy during an initial phase followed by the use of the adaptive strategy. We demonstrate the impact of adaptive training by searching for molecules that optimize both heuristic metrics, drug-likeness and synthesizability, as well as predicted protein binding affinity from a surrogate model. Our results show that the adaptive strategy provides a significant improvement in fitness optimization compared to the fixed pre-trained model, empowering the application of language models to molecular design tasks.
Chemical classification program synthesis using generative artificial intelligence
Accurately classifying chemical structures is essential for cheminformatics and bioinformatics, including tasks such as identifying bioactive compounds of interest, screening molecules for toxicity to humans, finding non-organic compounds with desirable material properties, or organizing large chemical libraries for drug discovery or environmental monitoring. However, manual classification is labor-intensive and difficult to scale to large chemical databases. Existing automated approaches either rely on manually constructed classification rules, or are deep learning methods that lack explainability. This work presents an approach that uses generative artificial intelligence to automatically write chemical classifier programs for classes in the Chemical Entities of Biological Interest (ChEBI) database. These programs can be used for efficient deterministic run-time classification of SMILES structures, with natural language explanations. The programs themselves constitute an explainable computable ontological model of chemical class nomenclature, which we call the ChEBI Chemical Class Program Ontology (C3PO). We validated our approach against the ChEBI database, and compared our results against deep learning models and a naive SMARTS pattern based classifier. C3PO outperforms the naive classifier, but does not reach the performance of state of the art deep learning methods. However, C3PO has a number of strengths that complement deep learning methods, including explainability and reduced data dependence. C3PO can be used alongside deep learning classifiers to provide an explanation of the classification, where both methods agree. The programs can be used as part of the ontology development process, and iteratively refined by expert human curators.
Genomic Language model for Annotation of Repetitive Elements (GLARE) v1.0
GLARE (Genomic Language model for Annotation of Repetitive Elements) is a tool that classifies transposable elements (TEs)—the mobile, repetitive DNA sequences that make up large fractions of eukaryotic genomes. GLARE fine-tunes the NTv3-650M genomic language model on a harmonized collection of curated TE sequences from the PanTEon and Repbase reference databases, assigning each input sequence to one of 11 orders and 32 superfamilies in a Wicker-compatible taxonomy. Features. From nucleotide FASTA input, GLARE outputs per-sequence predictions, class summaries, composition figures, and an annotated FASTA. It provides calibrated confidence scores with optional abstention and runs on CPU or GPU. Uses. GLARE serves as a classification component in genome-annotation pipelines, downstream of TE discovery, supporting genome annotation and comparative and evolutionary genomics. Advantages. GLARE is the first repeat-element classifier to leverage a pretrained genomic language model. Combined with multi-database training, this approach outperformed all nine classifiers in the PanTEon benchmark, generalized better to unseen taxonomic clades, and remained robust to sequence orientation—a common failure mode of existing tools.