Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “attention mechanism”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Graph-Based Attention Mechanisms for Solving the AC Optimal Power Flow Problem in Electrical Power Networks

With the increasing complexity and data availability in modern power systems, learning-based approaches to AC Optimal Power Flow (AC OPF) have garnered significant attention. In particular, the structure of smart grids lends itself naturally to graph-based representations, where Graph Neural Networks (GNNs) can capture spatial and relational dependencies. This paper investigates attention-based GNN architectures tailored to heterogeneous graph representations of electric grids. We evaluate two major paradigms: relational attention, which distinguishes between edge types during message passing, and meta-path attention, which captures high-level semantics through multi-hop, typed paths. Using a large corpus of public AC OPF scenarios, we benchmark representative models of each type of attention. Our results demonstrate the benefits of heterogeneous attention-based models in accurately capturing grid dynamics; heterogeneous attention models achieve superior performance in both standard and perturbed settings. The findings highlight the importance of semantic-aware architectures for improving prediction robustness and interpretability in power system applications.

Trigui, Ali [Qubit Engineering Inc.]↗
DOE OSTI2025

Enhancing Unknown Waveform Detection by Learning Intra and Inter-domain Dependencies with Advanced Attention Fusion Mechanisms

Detection of unknown waveforms in mission-critical communications is a crucial area of interest for the Department of Energy (DoE). Traditional methods and recent deep learning-based approaches often assume that the training set includes all possible classes, which is impractical for detecting new waveforms. This limitation gives rise to the problem of open-set recognition (OSR), which involves correctly identifying known classes while detecting and rejecting unknown or unseen classes. To address this limitation, we propose a novel dual-domain complex-valued neural architecture that jointly processes time-domain and frequency-domain signal representations using transformer mechanisms. A transformer model is a deep learning architecture that uses self-attention mechanisms to process and learn relationships in sequential data. Our model employs a cosine similarity loss to extract domain-specific features and incorporates a transformer architecture in the latent space to weigh the importance of different features from the time and frequency domains. The transformer layer includes stacked self-attention and cross-attention modules to learn intra-domain and inter-domain dependencies, creating a more holistic signal representation. An attention-based fusion module intelligently combines the time and frequency-domain features using multi-head attention, enabling the network to learn the optimal feature for each domain in each input signal. Quantitative results demonstrate the impact of these architectural choices on overall performance, showing significant improvement after incorporating self and cross-attention modules and using complex attention fusion over simple weighted fusion. Our ongoing work will focus on addressing the limitations of threshold-based OSR methods by developing a novel generative framework that integrates a conditional diffusion probabilistic model (DPM). DPM is a generative framework that learns to synthesize complex data by reversing a gradual noising process using a neural network trained to denoise step-by-step. Our goal is to leverage the inherent strengths of DPMs for identifying unknown signals more robustly. One primary advantage of using a DPM is its ability to provide a more reliable anomaly score based on the model's reconstruction error, rather than relying solely on classifier confidence. Additionally, the iterative denoising process of DPMs makes this approach naturally resilient to low Signal-to-Noise Ratio (SNR) conditions, where traditional methods often fail. By implementing this generative framework, we aim to enhance the model's capability to accurately detect unknown waveforms and maintain performance in challenging environments.

99 - GENERAL AND MISCELLANEOUS↗
DOE OSTI2025

Deformable phrase level attention: A flexible approach for improving AI based medical coding

Objective: Improving the AI-driven automated medical encoding of clinical text plays a vital role in gathering information on the occurrence of diseases to improve population-level health. This work presents a novel attention mechanism designed to enhance text classification models and ensure appropriate classification of medical concepts in unstructured electronic health records. Materials and Methods: We developed a deformable, phrase-level attention mechanism to identify important lexical word-level and contextual phrase-level information from clinical text documents. We evaluated conventional and transformer-based deep learning models that we extended with our attention mechanism on the extraction of critical cancer information (e.g., site, subsite, laterality, histology, behavior) from 629,908 electronic pathology reports and on the automated medical encoding of 52,722 hospital discharge summaries. Results: Transformer-based models with the deformable, phrase-level attention mechanism achieved the best performance on the extraction of critical cancer information from pathology reports. Conventional- and transformer-based models show similar or better performance than their baseline counterparts on the automated medical encoding of clinical documents. Discussion: The addition of phrase-level information allowed models extended with our proposed method to outperform standard word-level attention. Our method showed favorable properties for the real-world application in terms of model robustness and phenotyping. These results indicate that our method is promising for automated data harmonization for common data models. Conclusion: This work proposes a novel deformable, phrase-level attention mechanism that enhances text classification models in the extraction of medical concepts from clinical text documents. We demonstrate strong performances on two clinical text datasets and showcase real-world deployability of our method.

Automated medical encoding↗
DOE OSTI2025

CAMFeND: Credibility-Aware Multimodal Fake News Detection with Rotational Attention

In the evolving digital landscape, fake news is a significant challenge, influencing public perception and decision-making. Traditional detection approaches focus on single-modal data or simple multimodal fusion, often overlooking deeper interactions and news credibility. We propose a novel model addressing these limitations by introducing rotational attention and news domain information as a feature. Unlike static attention mechanisms, our rotational attention dynamically shifts query, key, and value roles across text and image inputs, enabling richer cross-modal interaction. Incorporating news domain information further enhances the model’s reliability by associating news posts with top domains extracted from Google search results, reducing false detections. This approach assesses both the content and the broader web context in which the news is discussed. Our model outperforms existing state-of-the-art methods by providing deeper, layered multimodal integration and domain information analysis, resulting in a more robust and adaptive fake news detection system.

Gupta, Nidhi↗
DOE OSTI2025

Towards Content Authenticity: Multimodal Fake News Detection and AI-Generated Text Identification

In today’s digital world, the spread of fake news and the rise of AI-generated text have become major threats to content authenticity and public trust. This thesis addresses both challenges through two complementary research directions: detecting fake news using multimodal features, and identifying AI-generated text using semantic and structural reasoning. The first part of the work focuses on fake news detection by introducing a novel model that combines text and image features through a unique rotational attention mechanism. Unlike traditional attention methods, this approach rotates the roles of query, key, and value across modalities to capture deeper interactions. Additionally, the model incorporates external domain information by linking news posts to top-ranked websites from Google search results, which helps assess the credibility of content based on its broader web context. This results in a more reliable and accurate fake news detection system that outperforms existing state-of-the-art methods. The second part presents SGG-ATD, a new framework for detecting AI-generated text. It uses masked language modeling to measure sentence coherence, followed by constructing a graph where keywords—both original and predicted—are connected based on semantic and contextual similarity. A Graph Convolutional Network (GCN) is then used to learn structural relationships within the text for final classification. Experimental results demonstrate that SGG-ATD achieves high F1-scores and consistently outperforms strong baselines. This method contributes to robust AI text detection, supporting accountability and resilience against AI-driven misinformation.

Gupta, Nidhi↗
DOE OSTI2026

When does global attention help: a unified empirical study on atomistic graph learning

Graph neural networks (GNNs) are widely used as surrogates for costly experiments and first-principles simulations to study the behavior of compounds at atomistic scale, and their architectural complexity is constantly increasing to enable the modeling of complex physics. While most recent GNNs combine more traditional message passing neural networks (MPNNs) layers to model short-range interactions with more advanced graph transformers (GTs) with global attention mechanisms to model long-range interactions, it is still unclear when global attention mechanisms provide real benefits over well-tuned MPNN layers due to inconsistent implementations, features, or hyperparameter tuning. We introduce the first unified, reproducible benchmarking framework–built on HydraGNN–that enables seamless switching among four controlled model classes: MPNN, MPNN with chemistry/topology encoders, GPS-style hybrids of MPNN with global attention, and fully fused localglobal models with encoders. Using seven diverse open-source datasets for benchmarking across regression and classification tasks, we systematically isolate the contributions of message passing, global attention, and encoder-based feature augmentation. Our study shows that encoder-augmented MPNNs form a robust baseline, while fused localglobal models yield the clearest benefits for properties governed by long-range interaction effects. We further quantify the accuracycompute trade-offs of attention, reporting its overhead in memory. Together, these results establish the first controlled evaluation of global attention in atomistic graph learning and provide a reproducible testbed for future model development.

Equivariant graph neural networks↗
DOE OSTI2025

Attention-based functional-group coarse-graining: a deep learning framework for molecular prediction and design

Machine learning (ML) offers considerable promise for the design of new molecules and materials. In real-world applications, the design problem is often domain-specific, and suffers from insufficient data, particularly labeled data, for ML training. In this study, we report a data-efficient, deep-learning framework for molecular discovery that integrates a coarse-grained functional-group representation with a self-attention mechanism to capture intricate chemical interactions. Our approach exploits group-contribution concepts to create a graph-based intermediate representation of molecules, serving as a low-dimensional embedding that substantially reduces the data demands typically required for training. Using a self-attention mechanism to learn the subtle but highly relevant chemical context of functional groups, the method proposed here consistently outperforms existing approaches for predictions of multiple thermophysical properties. In a case study focused on adhesive polymer monomers, we train on a limited dataset comprising only 6,000 unlabeled and 600 labeled monomers. The resulting chemistry prediction model achieves over 92% accuracy in forecasting properties directly from SMILES strings, exceeding the performance of current state-of-the-art techniques. Furthermore, the latent molecular embedding is invertible, enabling the design pipeline to automatically generate new monomers from the learned chemical subspace. We illustrate this functionality by targeting several properties, including high and low glass transition temperatures (Tg), and demonstrate that our model can identify new candidates with values that surpass those in the training set. The ease with which the proposed framework navigates both chemical diversity and data scarcity offers a promising route to accelerate and broaden the search for functional materials.

Han, Ming [Univ. of Chicago, IL (United States)]↗
DOE OSTI2025

RingX: Scalable Parallel Attention for Long-Context Learning on HPC

The attention mechanism has become foundational for remarkable AI breakthroughs since the introduction of the Transformer, driving the demand for increasingly longer context to power frontier models such as large-scale reasoning language models and high-resolution image/video generators. However, its quadratic computational and memory complexities present substantial challenges. Current state-of-the-art parallel attention methods, such as ring attention, are widely adopted for long-context training but utilize a point-to-point communication strategy that fails to fully exploit the capabilities of modern HPC network architectures. In this work, we propose ringX, a scalable family of parallel attention methods optimized explicitly for HPC systems. By enhancing workload partitioning, refining communication patterns, and improving load balancing, ringX achieves up to 3.4 × speedup compared to conventional ring attention on the Frontier supercomputer. Optimized for both bi-directional and causal attention mechanisms, ringX demonstrates its effectiveness through training benchmarks of a Vision Transformer (ViT) on a climate dataset and a Generative Pre-Trained Transformer (GPT) model, Llama3 8B. Our method attains an end-to-end training speedup of approximately 1.5 × in both scenarios. To our knowledge, the achieved 38% model FLOPs utilization (MFU) for training Llama3 8B with a 1M-token sequence length on 4,096 GPUs represents one of the highest training efficiencies reported for long-context learning on HPC systems. Our code implementation is available at https://github.com/jqyin/ringX-attention.

Yin, Junqi [ORNL] (ORCID:0000000338435520)↗
DOE OSTI2025

A Dynamic Hierarchical Attention Framework for Multimodal Malware Detection

The increasing use of Android in the worldwide mobile ecosystem has come along with a significant increase in advanced malware, highlighting the critical necessity for efficient, scalable, and adaptable detection systems. Despite recent advancements in machine learning improving malware detection, the majority of current solutions are limited to one, two, or three data modalities, hence neglecting the comprehensive behavioral spectrum of contemporary multi-vector threats. This thesis presents the first comprehensive multimodal framework for Android malware detection, which combines textual, time-series (temporal), graph-based (structural), and visual information using an innovative hierarchical attention mechanism and Dynamic Fusion Controller (DFC). Our methodology consistently classifies and processes modalities as either sequential or structural, facilitating content-adaptive weighting and resilient cross-modal representation learning. We advance the implementation of cutting-edge time series techniques, such as MiniRocket, for malware detection, hence creating new opportunities for temporal analysis in cybersecurity. Comprehensive experimental assessment shows that our framework performs exceptionally well, with 99.46% classification accuracy and 97.15% detection accuracy, significantly outperforming existing approaches through effective multimodal integration and hierarchical attention mechanisms.

Nazmin, Tamanna↗
DOE OSTI2025

ATTNChecker: Highly-Optimized Fault Tolerant Attention for Large Language Model Training

Large Language Models (LLMs) have demonstrated remarkable performance in various natural language processing tasks. However, the training of these models is computationally intensive and susceptible to faults, particularly in the attention mechanism, which is a critical component of transformer-based LLMs. In this paper, we investigate the impact of faults on LLM training, focusing on INF, NaN, and near-INF values in the computation results with systematic fault injection experiments. We observe the propagation patterns of these errors, which can trigger non-trainable states in the model and disrupt training, forcing the procedure to load from checkpoints. To mitigate the impact of these faults, we propose ATTNChecker, the first Algorithm-Based Fault Tolerance (ABFT) technique tailored for the attention mechanism in LLMs. ATTNChecker is designed based on fault propagation patterns of LLM and incorporates performance optimization to adapt to both system reliability and model vulnerability while providing lightweight protection for fast LLM training. Evaluations on four LLMs show that ATTNChecker on average incurs on average 7% overhead on training while detecting and correcting all extreme errors. Compared with the state-of-the-art checkpoint/restore approach, ATTNChecker reduces recovery overhead by up to 49×.

Liang, Yuhang [University of Alabama - Birmingham]↗
DOE OSTI2026

Hybrid Quantum–Classical Graph Transformers for Efficient Sentiment Analysis

Quantum Machine Learning (QML) offers a promising paradigm that leverages quantum computing principles to develop efficient and expressive models for learning from complex and structured data. Recent advances in natural language processing (NLP) and artificial intelligence (AI) have demonstrated capabilities in understanding, generating, and reasoning over linguistic and multimodal information. In this work, we present the Quantum Graph Transformer (QGT), a hybrid quantum–classical architecture that extends graph transformer capabilities through quantum self-attention. The QGT models variable-length sentences as token graphs, where both the embedding encoding and the self-attention mechanisms are implemented using parameterized quantum circuits (PQCs), enabling efficient contextual learning with significantly fewer trainable parameters. We train QGT using both fully connected and 𝑘 -nearest-neighbor graph structures and evaluate it on five benchmark sentiment-classification datasets. Experimental results show that QGT consistently achieves higher or comparable accuracy to existing quantum NLP models and outperforms a Classical Graph Transformer (CGT) baseline with identical architecture, achieving 29.4 × fewer parameters while requiring 3–5 × fewer samples to reach comparable performance. These findings highlight the potential of graph-based quantum models as scalable and data-efficient architectures for natural language understanding.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

River Dissolved Oxygen Prediction Using Machine Learning Models and Wireless Sensor Measurements

Simultaneous flooding&heat and droughts&heat events can potentially destabilize hydro-meteorological conditions to deteriorate the water quality of Neches River. Machine learning (ML) models utilizing wireless sensor measurements have been applied to predict water quality and optimize various water management strategies. This study aims to develop ML models to predict dissolved oxygen (DO) prediction under various hydro-meteorological conditions and enhance water management decision-making. Wireless sensor measurements of DO, water temperature, sample depth, conductivity, turbidity, and pH, along with discharge from the United States Geological Survey stations, are collected for model inputs at the Pine Island Bayou C749 station (PIB-C749) and Neches River Saltwater Barrier (SWB). Multilayer perceptron neural networks, recurrent neural networks, long short-term memory (LSTM), and bidirectional LSTM (BiLSTM) with and without attention mechanism (AT) are tested to determine the best model, which is applied the rolling forecast method to predict 14-day DO. Traditional and recurrent transfer learning (TL and RTL) methods are adopted to overcome insufficient data at the SWB. The input feature importance analysis using the integrated gradients (IG) algorithm is applied to determine dominant inputs. The results show LSTM-based models are capable handling long sequential data. AT-BiLSTM and RTL-LSTM demonstrate the best performance at the PIB-C749 (RMSE=0.054) and the SWB (RMSE=0.028), respectively. TL and RTL methods significantly improve model performance at the SWB. DO, temperature, and pH show higher importance, consistent with hydrodynamics and water chemistry. Both best models are applied to predict 14-day DO and demonstrate reasonable performance for decision-making. Hydro-meteorological conditions of 2017 flood and 2012 drought events are simulated and reveal that possible hypoxia occurs after flooding due to increasing temperature and turbidity, and DO concentration decreases significantly under heat and drought conditions. In conclusion, LSTM-based models utilizing wireless sensor data can be a timely and effective approach to make appropriate decisions on water resource management.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

HARMONY: Large-Scale Architecture Search for Efficient Hybrid Language Models

As large language models scale to trillions of parameters, their computational and memory requirements present critical challenges for efficient training and deployment. While Mixture of Experts (MoE) architectures enable efficient scaling through sparse parameter activation, and state-space models like Mamba offer linear-time complexity, principled methods for combining these paradigms remain undeveloped. We introduce HARMONY (Hybrid Architecture Research for Mamba, Optimized with Neural efficiencY), a multi-objective evolutionary neural architecture search framework for discovering efficient hybrid language models that integrate Transformer attention mechanisms, Mixture-of-Experts routing, and Mamba state-space components. Through large-scale distributed search using 16,384 MI250X GPUs on the Frontier supercomputer, HARMONY explores a comprehensive design space encompassing six attention variants (MHA, MQA, GQA, MLA, SWA, and Mamba-2), variable MoE configurations with both routed and shared experts, and extensive Mamba hyperparameters. Our framework discovers heterogeneous architectures that balance training performance with computational efficiency through multi-objective optimization incorporating latency penalties and fitness-based selection. Analysis of discovered architectures reveals that optimal hybrid designs favor heterogeneous component mixing rather than homogeneous patterns, with Mamba-2 and Multi-Head Latent Attention (MLA) emerging as preferred mechanisms. Discovered architectures demonstrate superior training efficiency: our best configuration achieves a final perplexity of 1.0874 with 2.38B parameters while processing 4,320 tokens/second, outperforming significantly larger manually designed models. Full-scale evaluation shows HARMONY's top architectures achieve better loss trajectories than equivalently-sized models using state-of-the-art configurations including Mixtral, Jamba, and Samba. Additionally, we demonstrate 91% weak scaling efficiency when training discovered 36B-parameter models across 1,024 GPUs. HARMONY is released as an open framework with comprehensive tools for building and training hybrid models using expert-data-pipeline parallelism, democratizing access to automated architecture design for next-generation language models.

Herron, Emily [ORNL] (ORCID:0000000273008172)↗
DOE OSTI2025

Attention to quantum complexity

The imminent era of error-corrected quantum computing demands robust methods to characterize quantum state complexity from limited, noisy measurements. We introduce the Quantum Attention Network (QuAN), a classical artificial intelligence (AI) framework leveraging attention mechanisms tailored for learning quantum complexity. Inspired by large language models, QuAN treats measurement snapshots as tokens while respecting permutation invariance. Combined with our parameter-efficient miniset self-attention block, this enables QuAN to access high-order moments of bit-string distributions and preferentially attend to less noisy snapshots. We test QuAN across three quantum simulation settings: driven hard-core Bose-Hubbard model, random quantum circuits, and toric code under coherent and incoherent noise. QuAN directly learns entanglement and state complexity growth from experimental computational basis measurements, including complexity growth in random circuits from noisy data. In regimes inaccessible to existing theory, QuAN unveils the complete phase diagram for noisy toric code data as a function of both noise types, highlighting AI’s transformative potential for assisting quantum hardware.

Kim, Hyejin [Cornell Univ., Ithaca, NY (United Sta↗
DOE OSTI2025

Adaptive Graph Learning with Transformer for Multi-Reservoir Inflow Prediction

Reservoir inflow prediction is crucial for water resource management, yet existing approaches mainly focus on single-reservoir models that ignore spatial dependencies among interconnected reservoirs. We introduce AdaTrip as an adaptive, time-varying graph learning framework for multi-reservoir inflow forecasting. AdaTrip constructs dynamic graphs where reservoirs are nodes with directed edges reflecting hydrological connections, employing attention mechanisms to automatically identify crucial spatial and temporal dependencies. Evaluation on thirty reservoirs in the Upper Colorado River Basin demonstrates superiority over existing baselines, with improved performance for reservoirs with limited records through parameter sharing. Additionally, AdaTrip provides interpretable attention maps at edge and time-step levels, offering insights into hydrological controls to support operational decision-making. Our code is available at https://github.com/humphreyhuu/AdaTrip.

Hu, Pengfei [ORNL] (ORCID:0009000367130950)↗
DOE OSTI2025

Why Is Attention Sparse In Particle Transformer?

Transformer-based models have achieved state-of-the-art performance in jet tagging at the CERN Large Hadron Collider (LHC), with the Particle Transformer (ParT) representing a leading example of such models. A striking feature of ParT is its sparse, nearly binary, attention structure, raising questions about the origin of this behavior and whether it encodes physically meaningful correlations. In this work, we investigate the source of ParT's sparse attention by comparing models trained on multiple benchmark datasets and examine the relative contributions of the attention term and the physics-inspired interaction matrix before softmax. We find that binary sparsity arises primarily from the attention mechanism itself, with the interaction matrix playing a secondary role. Moreove, we show that ParT is able to identify key jet substructure elements, such as leptons in semileptonic top decays, even without explicit particle identification inputs. These results provide new insight into the interpretability of transformer-based jet taggers and clarify the conditions under which sparse attention patterns emerge in ParT.

Legge, Timothy [UC, San Diego]↗