Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “self-supervised”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Accelerated Over-The-Air Neural Receiver Training Using Self-Contrastive Learning

Self-contrastive learning (SCL), a self-supervised learning method, has been shown to improve image and signal classifier accuracies and reduce the training time for neural communications receivers. In particular, prior work has shown that SCL applied as a pre-training step can improve simulated performance of OFDM in 3GPP TDL channel models by reducing the training time of the downstream classification task (demodulation and demapping). In this work a practical implementation demonstrating SCL pre-training using software defined radios (SDRs) is proposed.

Cooke, Corey [ORNL] (ORCID:0000000234263672)↗
DOE OSTI2025

Vision Foundation Models in Remote Sensing: A survey

Artificial intelligence (AI) technologies have profoundly transformed the field of remote sensing (RS), revolutionizing data collection, processing, and analysis. Traditionally reliant on manual interpretation and task-specific models, RS research has been significantly enhanced by the advent of foundation models (FMs)—large-scale pretrained AI models capable of performing a wide array of tasks with unprecedented accuracy and efficiency. This article provides a comprehensive survey of FMs in the RS domain. We categorize these models based on their architectures, pretraining datasets, and methodologies. Through detailed performance comparisons, we highlight emerging trends and the significant advancements achieved by those FMs. Additionally, we discuss technical challenges, practical implications, and future research directions, addressing the need for high-quality data, computational resources, and improved model generalization. Our research also finds that pretraining methods, particularly self-supervised learning (SSL) techniques like contrastive learning (CL) and masked autoencoders (MAEs), remarkably enhance the performance and robustness of FMs. This survey aims to serve as a resource for researchers and practitioners by providing a panorama of advances and promising pathways for the continued development and application of FMs in RS.

data models↗
DOE OSTI2025

CareWELL: Multimodal Region Representation Learning with Spatial Contexts for Urban Health

Rapid urbanization affects living environments by intensifying exposure to air pollution, heat, noise, and urban dynamics, which together contribute to uneven health outcomes across neighborhoods. For instance, cardiovascular, respiratory, and mental health conditions are each influenced by distinct exposures such as air pollution, extreme temperatures, or limited access to green space. These heterogeneous patterns require understanding the characteristics of geographic regions in order to explain why urban health risks vary across urban areas. Recent work in self-supervised region representation learning provides a promising way to model such characteristics from multimodal geospatial data. However, existing methods face two major limitations: (i) they often depend on non-public datasets, limiting reproducibility and applicability, and (ii) their generic pretraining objectives overlook health-relevant determinants, including temporal variability in environmental exposures and inequalities in social conditions. To address these gaps, we propose Context-Aware Region rEpresentation with Weather, Environment, and Location Learning (CareWELL). CareWELL leverages large language models to encode seasonal variability in weather, employs contrastive learning to align geo-coordinate and weather representations, and introduces a context-aware objective that integrates socio-demographic factors while preserving spatial correlations. We evaluate CareWELL by predicting six urban health outcomes in Manhattan, New York City, and demonstrate that CareWELL consistently outperforms state-of-the-art baselines as well as a traditional spatial computing method. These results suggest the importance of context-aware pretraining objectives for learning health-relevant region representations.

Namgung, Min [ORNL]↗
DOE OSTI2026

Emerging Flexible Designs for Geospatial Multimodal Foundation Models

Foundation models are rapidly transforming Earth observation by enabling scalable pretraining across diverse unlabeled geospatial modalities. However, their architectural diversity—ranging from encoder-only to encoder-decoder and masked autoencoding paradigms—makes it challenging to assess performance trade-offs in a consistent manner. In this work, we present an apples-to-apples comparison of leading FM architectures designed for geospatial multimodal reasoning, with a particular focus on flexibility across varied spectral band configurations. We standardize pretraining using identical self-supervised learning objectives and training datasets, and evaluate all models under consistent parameterization on the GEOBench benchmark across classification and segmentation tasks. Our results offer new insights into the design trade-offs between model flexibility, modality alignment, and downstream task performance. By highlighting architectural strengths and limitations under controlled conditions, this study provides practical guidance for building next-generation geospatial foundation models capable of robust multimodal reasoning.

Ambrozio Dias, Philipe [ORNL] (ORCID:0000000194277↗
DOE OSTI2025

A knowledge-informed large language model framework for U.S. nuclear power plant shutdown initiating event classification for probabilistic risk assessment

Identifying and classifying shutdown initiating events (SDIEs) is critical for developing shutdown probabilistic risk assessment for nuclear power plants. Existing computational approaches cannot achieve satisfactory performance due to the challenges of unavailable large, labeled datasets, imbalanced event types, and label noise. To address these challenges, we propose a hybrid pipeline that integrates a knowledge-informed machine learning model to prescreen non-SDIEs and a large language model (LLM) to classify SDIEs into four types. In the prescreening stage, we proposed a set of 44 SDIE text patterns that consist of the most salient keywords and phrases from six SDIE types. Text vectorization based on the SDIE patterns generates feature vectors that are highly separable by using a simple binary classifier. The second stage builds Bidirectional Encoder Representations from Transformers (BERT)-based LLM, which learns generic English language representations from self-supervised pretraining on a large dataset and adapts to SDIE classification by fine-tuning it on an SDIE dataset. The proposed approaches are evaluated on a dataset with 10,928 events using precision, recall ratio, F 1 score, and average accuracy. In conclusion, the results demonstrate that the prescreening stage can exclude more than 97% non-SDIEs, and the LLM achieves an average accuracy of 95.1% for SDIE classification.

99 - GENERAL AND MISCELLANEOUS↗
DOE OSTI2025

The role of chromatin state in intron retention: A case study in leveraging large scale deep learning models

Complex deep learning models trained on very large datasets have become key enabling tools for current research in natural language processing and computer vision. By providing pre-trained models that can be fine-tuned for specific applications, they enable researchers to create accurate models with minimal effort and computational resources. Large scale genomics deep learning models come in two flavors: the first are large language models of DNA sequences trained in a self-supervised fashion, similar to the corresponding natural language models; the second are supervised learning models that leverage large scale genomics datasets from ENCODE and other sources. We argue that these models are the equivalent of foundation models in natural language processing in their utility, as they encode within them chromatin state in its different aspects, providing useful representations that allow quick deployment of accurate models of gene regulation. We demonstrate this premise by leveraging the recently created Sei model to develop simple, interpretable models of intron retention, and demonstrate their advantage over models based on the DNA language model DNABERT-2. Our work also demonstrates the impact of chromatin state on the regulation of intron retention. Using representations learned by Sei, our model is able to discover the involvement of transcription factors and chromatin marks in regulating intron retention, providing better accuracy than a recently published custom model developed for this purpose.

Biochemistry & Molecular Biology↗
DOE OSTI2025

Machine Learning for Anomaly Detection in Neural Network Security and SRF Cavities

This dissertation explores the development and deployment of machine learning approaches to address critical challenges in anomaly detection across two distinct domains: neural network security in federated learning settings and cavity behavior analysis in particle accelerator operations at Jefferson Lab in Newport News, Virginia. Anomaly detection identifies deviations from expected patterns, safeguarding systems in cybersecurity, industry, and research against malicious activities and failures. This dissertation demonstrates how our machine learning approaches enhance detection accuracy and efficiency in both neural network security and industrial applications. First, we investigate vulnerabilities in deep neural networks deployed in federated learning. Although federated learning preserves user privacy by training models locally, it remains vulnerable to backdoor attacks, in which malicious participants embed hidden triggers that induce targeted misbehavior. We propose a self-supervised contrastive learning framework to detect and mitigate such backdoor attacks. In our experiments, this method achieves higher detection accuracy and lower false positive rates than existing defenses, while operating without access to local model updates or original training data and thus preserving the privacy guarantees of the federated setting. Second, we address the operational reliability of superconducting radio-frequency (SRF) cavities at the Continuous Electron Beam Accelerator Facility (CEBAF). Our research leverages an unsupervised learning approach, combined with Principal Component Analysis (PCA) and k-means clustering, to identify anomalous behaviors in SRF cavities. Our method detects subtle anomalous behavior by analyzing SRF signal data. This knowledge allows for the early detection and resolution of potential faults, significantly improving the efficiency and reliability of operations. Third, we extend these insights to time-series anomaly detection more broadly. We design a contrastive-learning based model tailored to increasingly dynamic environments and academic research. This model improves detection accuracy in settings that require real-time monitoring and predictive maintenance. Our research underscores the broader applicability and impact of advanced machine learning techniques in anomaly detection. By extracting meaningful patterns from complex data, machine learning can significantly enhance security in distributed neural networks and improve the efficiency of particle accelerator operations. This dissertation serves as a stepping stone for future investigations into the vast possibilities of anomaly detection, inspiring further exploration and development of machine learning techniques in this field.

Ferguson, Hal [Old Dominion University]↗
DOE OSTI2026

A Semi-Supervised Learning Method for the Identification of Bad Exposures in Large Imaging Surveys

As the data volume of astronomical imaging surveys rapidly increases, traditional methods for image anomaly detection, such as visual inspection by human experts, are becoming impractical. We introduce a machine-learning-based approach to detect poor-quality exposures in large imaging surveys, with a focus on the DECam Legacy Survey (DECaLS) in regions of low extinction (i.e., E ( B − V ) < 0.04 ). Our semi-supervised pipeline integrates a vision transformer (ViT), trained via self-supervised learning (SSL), with a k-Nearest Neighbor (kNN) classifier. We train and validate our pipeline using a small set of labeled exposures observed by surveys with the Dark Energy Camera (DECam). A clustering-space analysis of where our pipeline places images labeled in good and bad categories suggests that our approach can efficiently and accurately determine the quality of exposures. Applied to new imaging being reduced for DECaLS Data Release 11, our pipeline identifies 780 problematic exposures, which we subsequently verify through visual inspection. Being highly efficient and adaptable, our method offers a scalable solution for quality control in other large imaging surveys.

Luo, Yufeng (ORCID:0000000246230683)↗
DOE OSTI2026

Data-driven particle dynamics: Structure-preserving coarse-graining for emergent behavior in non-equilibrium systems

Multiscale systems are ubiquitous in science and technology, but are notoriously challenging to simulate as short spatiotemporal scales must be appropriately linked to emergent bulk physics. When expensive high-dimensional dynamical systems are coarse-grained into low-dimensional models, the entropic loss of information leads to emergent physics which are dissipative, history-dependent, and stochastic. To machine learn coarse-grained dynamics from time-series observations of particle trajectories, we propose a framework using the metriplectic bracket formalism that preserves these properties by construction; most notably, the framework guarantees discrete notions of the first and second laws of thermodynamics, conservation of momentum, and a discrete fluctuation-dissipation balance crucial for capturing non-equilibrium statistics. We introduce the mathematical framework abstractly before specializing to a particle discretization. As labels are generally unavailable for entropic state variables, we introduce a novel self-supervised learning strategy to identify emergent structural variables. We validate the method on benchmark systems and demonstrate its utility on two challenging examples: (1) coarse-graining star polymers at challenging levels of coarse-graining while preserving non-equilibrium statistics, and (2) learning models from high-speed video of colloidal suspensions that capture coupling between local rearrangement events and emergent stochastic dynamics. We provide open-source implementations in both PyTorch and LAMMPS, enabling large-scale inference and extensibility to diverse particle-based systems.

Computational Engineering, Finance, and Science (c↗
NASA NTRS1980

Fitness variables and the lipid profile in United States astronauts

The study examines the relationship between several measures of fitness and the lipid profile in United States astronauts. Data were collected on 89 astronauts, previously selected (PSA) and newly selected (NSA), during their annual physical examinations. Several similarities were seen in the two groups. The PSA (mean age of 46.1) had a lower maximum oxygen capacity (41.7 ml kg/min vs. 47.5 ml kg/min); when adjusted for age, it was no different from the NSA (mean age 33.5). The PSA had similar body composition with 15.7% - lower than expected for age. The lipid profiles of the two groups were basically the same with the differences being a function of age. Compared to a normative population, the astronauts had similar cholesterols, lower triglycerides, and higher HDLs. The astronaut profiles were generally more favorable than the age-matched controls, which is felt to be a result of the self-supervised conditioning program and annual preventive medicine consultation and education.

Berry, M. A.↗
NASA NTRS2009

Complexity for Survival of Living Systems

A logical connection between the survivability of living systems and the complexity of their behavior (equivalently, mental complexity) has been established. This connection is an important intermediate result of continuing research on mathematical models that could constitute a unified representation of the evolution of both living and non-living systems. Earlier results of this research were reported in several prior NASA Tech Briefs articles, the two most relevant being Characteristics of Dynamics of Intelligent Systems (NPO- 21037), NASA Tech Briefs, Vol. 26, No. 12 (December 2002), page 48; and Self-Supervised Dynamical Systems (NPO- 30634) NASA Tech Briefs, Vol. 27, No. 3 (March 2003), page 72. As used here, living systems is synonymous with active systems and intelligent systems. The quoted terms can signify artificial agents (e.g., suitably programmed computers) or natural biological systems ranging from single-cell organisms at one extreme to the whole of human society at the other extreme. One of the requirements that must be satisfied in mathematical modeling of living systems is reconciliation of evolution of life with the second law of thermodynamics. In the approach followed in this research, this reconciliation is effected by means of a model, inspired partly by quantum mechanics, in which the quantum potential is replaced with an information potential. The model captures the most fundamental property of life - the ability to evolve from disorder to order without any external interference. The model incorporates the equations of classical dynamics, including Newton s equations of motion and equations for random components caused by uncertainties in initial conditions and by Langevin forces. The equations of classical dynamics are coupled with corresponding Liouville or Fokker-Planck equations that describe the evolutions of probability densities that represent the uncertainties. The coupling is effected by fictitious information-based forces that are gradients of the information potential, which, in turn, is a function of the probability densities. The probability densities are associated with mental images both self-image and nonself images (images of external objects that can include other agents). The evolution of the probability densities represents mental dynamics. Then the interaction between the physical and metal aspects of behavior is implemented by feedback from mental to motor dynamics, as represented by the aforementioned fictitious forces. The interaction of a system with its self and nonself images affords unlimited capacity for increase of complexity. There is a biological basis for this model of mental dynamics in the discovery of mirror neurons that learn by imitation. The levels of complexity attained by use of this model match those observed in living systems. To establish a mechanism for increasing the complexity of dynamics of an active system, the model enables exploitation of a chain of reflections exemplified by questions of the form, "What do you think that I think that you think...?" Mathematically, each level of reflection is represented in the form of an attractor performing the corresponding level of abstraction with more details removed from higher levels. The model can be used to describe the behaviors, not only of biological systems, but also of ecological, social, and economics ones.

Zak, Michail↗
NASA NTRS2009

NASA Tech Briefs, August 2009

Topics covered include: Aligning a Receiving Antenna Array to Reduce Interference; Collecting Ground Samples for Balloon-Borne Instruments; Tethered Pyrotechnic Apparatus for Acquiring a Ground Sample; Enhanced Video-Oculography System; Joint Carrier-Phase Synchronization and LDPC Decoding; Dual-Polarization, Sideband-Separating, Balanced Receiver for 1.5 THz Modular Battery Charge Controller; Efficient Multiplexer FPGA Block Structures Based on G4FETs; VLSI Microsystem for Rapid Bioinformatic Pattern Recognition; Low-Noise Amplifier for 100 to 180 GHz; Improved Fabrication of Ceramic Matrix Composite/Foam Core Integrated Structures; Inert Welding/Brazing Gas Filters and Dryers; Fabricating Copper Nanotubes by Electrodeposition; Reducing Aerodynamic Drag on Empty Open Cargo Vehicles; Rotary Percussive Auto-Gopher for Deep Drilling and Sampling; More About Reconfigurable Exploratory Robotic Vehicles; Thermostatic Valves Containing Silicone-Oil Actuators; Improving Heat Flux Performance of Flat Surface in Spray-Cooling Systems; Treating Fibrous Insulation to Reduce Thermal Conductivity; Silica-Aerogel Composites Opacified with La(sub0.7)Sr(sub0.3)MnO3; Cyclic Oxidation Behavior of CuCrAl Cold-Sprayed Coatings for Reusable Launch Vehicles; Ceramic Fiber Structures for Cryogenic Load-Bearing Applications; Elastomer Reinforced with Carbon Nanotubes; Biologically Inspired Purification and Dispersion of SWCNTs; A Technique for Adjusting Eigenfrequencies of WGM Resonators; Low-Pressure, Field-Ionizing Mass Spectrometer; Modifying Operating Cycles to Increase Stability in a LITS; Chamber for Simulating Martian and Terrestrial Environments; Algorithm for Detecting a Bright Spot in an Image; Extreme Programming: Maestro Style; Adaptive Behavior for Mobile Robots; Protocol for Communication Networking for Formation Flying; Planning Complex Sequences Using Compressed Representations; and Self-Supervised Learning of Terrain Traversability from Proprioceptive Sensors.

Source record↗
NASA NTRS2010

Modeling Common-Sense Decisions in Artificial Intelligence

A methodology has been conceived for efficient synthesis of dynamical models that simulate common-sense decision- making processes. This methodology is intended to contribute to the design of artificial-intelligence systems that could imitate human common-sense decision making or assist humans in making correct decisions in unanticipated circumstances. This methodology is a product of continuing research on mathematical models of the behaviors of single- and multi-agent systems known in biology, economics, and sociology, ranging from a single-cell organism at one extreme to the whole of human society at the other extreme. Earlier results of this research were reported in several prior NASA Tech Briefs articles, the three most recent and relevant being Characteristics of Dynamics of Intelligent Systems (NPO -21037), NASA Tech Briefs, Vol. 26, No. 12 (December 2002), page 48; Self-Supervised Dynamical Systems (NPO-30634), NASA Tech Briefs, Vol. 27, No. 3 (March 2003), page 72; and Complexity for Survival of Living Systems (NPO- 43302), NASA Tech Briefs, Vol. 33, No. 7 (July 2009), page 62. The methodology involves the concepts reported previously, albeit viewed from a different perspective. One of the main underlying ideas is to extend the application of physical first principles to the behaviors of living systems. Models of motor dynamics are used to simulate the observable behaviors of systems or objects of interest, and models of mental dynamics are used to represent the evolution of the corresponding knowledge bases. For a given system, the knowledge base is modeled in the form of probability distributions and the mental dynamics is represented by models of the evolution of the probability densities or, equivalently, models of flows of information. Autonomy is imparted to the decisionmaking process by feedback from mental to motor dynamics. This feedback replaces unavailable external information by information stored in the internal knowledge base. Representation of the dynamical models in a parameterized form reduces the task of common-sense-based decision making to a solution of the following hetero-associated-memory problem: store a set of m predetermined stochastic processes given by their probability distributions in such a way that when presented with an unexpected change in the form of an input out of the set of M inputs, the coupled motormental dynamics converges to the corresponding one of the m pre-assigned stochastic process, and a sample of this process represents the decision.

Zak, Michail↗
NASA NTRS2011

NASA Tech Briefs, May 2011

Topics covered include: 1) Method to Estimate the Dissolved Air Content in Hydraulic Fluid; 2) Method for Measuring Collimator-Pointing Sensitivity to Temperature Changes; 3) High-Temperature Thermometer Using Cr-Doped GdAlO3 Broadband Luminescence; 4)Metrology Arrangement for Measuring the Positions of Mirrors of a Submillimeter Telescope; 5) On-Wafer S-Parameter Measurements in the 325-508-GHz Band; 6) Reconfigurable Microwave Phase Delay Element for Frequency Reference and Phase-Shifter Applications; 7) High-Speed Isolation Board for Flight Hardware Testing; 8) High-Throughput, Adaptive FFT Architecture for FPGA-Based Spaceborne Data Processors; 9) 3D Orbit Visualization for Earth-Observing Missions; 10) MaROS: Web Visualization of Mars Orbiting and Landed Assets; 11) RAPID: Collaborative Commanding and Monitoring of Lunar Assets; 12) Image Segmentation, Registration, Compression, and Matching; 13) Image Calibration; 14) Rapid ISS Power Availability Simulator; 15) A Method of Strengthening Composite/Metal Joints; 16) Pre-Finishing of SiC for Optical Applications; 17) Optimization of Indium Bump Morphology for Improved Flip Chip Devices; 18) Measuring Moisture Levels in Graphite Epoxy Composite Sandwich Structures; 19) Marshall Convergent Spray Formulation Improvement for High Temperatures; 20) Real-Time Deposition Monitor for Ultrathin Conductive Films; 21) Optimized Li-Ion Electrolytes Containing Triphenyl Phosphate as a Flame-Retardant Additive; 22) Radiation-Resistant Hybrid Lotus Effect for Achieving Photoelectrocatalytic Self-Cleaning Anticontamination Coatings; 23) Improved, Low-Stress Economical Submerged Pipeline; 24) Optical Fiber Array Assemblies for Space Flight on the Lunar Reconnaissance Orbiter; 25) Local Leak Detection and Health Monitoring of Pressurized Tanks; 26) Dielectric Covered Planar Antennas at Submillimeter Wavelengths for Terahertz Imaging; 27) Automated Cryocooler Monitor and Control System; 28) Broadband Achromatic Phase Shifter for a Nulling Interferometer; 29) Super Dwarf Wheat for Growth in Confined Spaces; 30) Fine Guidance Sensing for Coronagraphic Observatories; 31) Single-Antenna Temperature- and Humidity-Sounding Microwave Receiver; 32) Multi-Wavelength, Multi-Beam, and Polarization-Sensitive Laser Transmitter for Surface Mapping; 33) Optical Communications Link to Airborne Transceiver; 34) Ascent Heating Thermal Analysis on Spacecraft Adaptor Fairings; 35) Entanglement in Self-Supervised Dynamics; 36) Prioritized LT Codes; 37) Fast Image Texture Classification Using Decision Trees; 38) Constraint Embedding Technique for Multibody System Dynamics; 39) Improved Systematic Pointing Error Model for the DSN Antennas; 40) Observability and Estimation of Distributed Space Systems via Local Information-Exchange Networks; 41) More-Accurate Model of Flows in Rocket Injectors; 42) In-Orbit Instrument-Pointing Calibration Using the Moon as a Target; 43) Reliability of Ceramic Column Grid Array Interconnect Packages Under Extreme Temperatures; 44) Six Degrees-of-Freedom Ascent Control for Small-Body Touch and Go; and 45) Optical-Path-Difference Linear Mechanism for the Panchromatic Fourier Transform Spectrometer.

Source record↗
NASA NTRS2011

Introduction to Physical Intelligence

A slight deviation from Newtonian dynamics can lead to new effects associated with the concept of physical intelligence. Non-Newtonian effects such as deviation from classical thermodynamic as well as quantum-like properties have been analyzed. A self-supervised (intelligent) particle that can escape from Brownian motion autonomously is introduced. Such a capability is due to a coupling of the particle governing equation with its own Liouville equation via an appropriate feedback. As a result, the governing equation is self-stabilized, and random oscillations are suppressed, while the Liouville equation takes the form of the Fokker-Planck equation with negative diffusion. Non- Newtonian properties of such a dynamical system as well as thermodynamical implications have been evaluated.

Zak, Michail↗
NASA NTRSDate not supplied

AI Foundation Models for Science: An Open Collaborative Initiative

Foundation Models (FMs), AI models designed to replace task-specific models, are increasingly being recognized for their versatility across numerous downstream applications. These models, trained using self-supervised techniques on any type of sequence data, circumvent the need for large annotated datasets, a major bottleneck in traditional AI model development. FMs can be applied to downstream tasks using few-shot learning and fine-tuning, significantly reducing the need for large labeled training datasets and computational resources. However, the development of FMs requires substantial resources, including access to data and compute power, expertise in the latest models, and specialized scientific knowledge for systematic evaluation. It is challenging for a single group to possess all these capabilities. To address this, NASA IMPACT has initiated an open collaborative effort, leveraging partnerships with the private sector and other groups within and outside NASA, to jointly build FMs. The overarching goal is to develop a consistent and collaborative approach to building FMs for high-value science datasets. This initiative has fostered collaboration within NASA and with external partners, including IBM Research, Clark University, DOE’s ORNL, ESA, and USGS. The effort focuses on identifying key datasets with a wide range of downstream applications, pretraining and building FMs using modified transformer architectures, evaluating compute infrastructure needs, and sharing models, pretraining and fine-tuning code, and data with the community. Furthermore, it aims to train the Earth science community to fine-tune these models for various downstream applications. Our initial effort resulted in the creation of a 100 million parameter HLS Geospatial Model within six months, which was released on HuggingFace. We are now expanding our scope to include data from weather and climate models and investigating multimodal models. We invite those interested in participating in this effort to join us by sharing their use cases, expertise, or data.

Rahul Ramachandran↗
DOE OSTI2024

Evaluating the Use of Foundational Chemical Language Models in Multimodal Graph Fusion

Rapid and accurate prediction of the physicochemical properties of molecules given their structures remains a key challenge in cheminformatics. Machine learning approaches offer high-throughput options, but the optimality of inductive biases and data representations are up for debate. For example, BERT-based masked language models (MLMs) can be trained in a self-supervised way on hundreds of millions to billions of readily available SMILES strings. Another option is graph neural networks (GNNs), which can operate directly on molecular structures. Yet, generating accurate molecular geometry is computationally expensive, leading to a relative scarcity in data compared to SMILES strings. It is attractive to combine these two paradigms by pre-training an LM on a large corpus of SMILES strings and embedding these representation into a geometric graph neural network. Despite the promise of such an approach, and contrary to previous studies, we find mixed results with the combination of the LMs and GNNs on several molecule datasets. In particular, we found evidence for improvement on the FreeSolv and QM7 benchmarks, but degraded performance on the ESOL, LIPO and QM9 datasets compared to a GNN baseline.

Francel, Collin [University of Alabama]↗
DOE OSTI2025

RINO: Renormalization Group Invariance with No Labels

A common challenge with supervised machine learning (ML) in high energy physics (HEP) is the reliance on simulations for labeled data, which can often mismodel the underlying collision or detector response. To help mitigate this problem of domain shift, we propose RINO (Renormalization Group Invariance with No Labels), a self-supervised learning approach that can instead pretrain models directly on collision data, learning embeddings invariant to renormalization group flow scales. In this work, we pretrain a transformer-based model on jets originating from quantum chromodynamic (QCD) interactions from the JetClass dataset, emulating real QCD-dominated experimental data, and then finetune on the JetNet dataset -- emulating simulations -- for the task of identifying jets originating from top quark decays. RINO demonstrates improved generalization from the JetNet training data to JetClass data compared to supervised training on JetNet from scratch, demonstrating the potential for RINO pretraining on real collision data followed by fine-tuning on small, high-quality MC datasets, to improve the robustness of ML models in HEP.

Hao, Zichun [Caltech] (ORCID:0000000256244907)↗