pnnl/Multimodal-Few-Shot
This project combines few-shot machine learning with a multimodal framework for electron microscope data quantification.
Doty, Christina↗
SEARCH · Engineering Papers
Results for “few shot learning”
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.
Development of Multimodal Few-Shot Analytics for Electron Micrographs
Recent advances in materials data analytics have provided new avenues for determining process-structure-property (PSP) linkages in a variety of materials. Machine learning techniques including few-shot learning have increased the efficiency of classifying microscopy images for the purposes of material characterization. Attempts at creating a multimodal approach can provide further improvements to current models and help extract more salient features from data. In this vein, raw spectrum data was taken to provide an additional modality to our current pyCHIP classifier. Modifications in segmentation also show potential in improving the accuracy of the pyCHIP classifier. Classifier output was analyzed using network graphs and unsupervised clustering algorithms such as spectral clustering to detect better segmentation methods than the current “chipping” approach. We suggest that the chip selection process can be automated in the future using a combination of these techniques to enable high-throughput analyses.
36 MATERIALS SCIENCE↗
QCalEval: Benchmarking Vision-Language Models for Quantum Calibration Plot Understanding
Quantum computing calibration depends on interpreting experimental data, and calibration plots provide the most universal human-readable representation for this task, yet no systematic evaluation exists of how well vision-language models (VLMs) interpret them. We introduce QCalEval, the first VLM benchmark for quantum calibration plots: 243 samples across 87 scenario types from 22 experiment families, spanning superconducting qubits and neutral atoms, evaluated on six question types in both zero-shot and in-context learning settings. The best general-purpose zero-shot model reaches a mean score of 72.3, and many open-weight models degrade under multi-image in-context learning, whereas frontier closed models improve substantially. A supervised fine-tuning ablation at the 9-billion-parameter scale shows that SFT improves zero-shot performance but cannot close the multimodal in-context learning gap. As a reference case study, we release NVIDIA Ising Calibration 1, an open-weight model based on Qwen3.5-35B-A3B that reaches 74.7 zero-shot average score.
Cao, Shuxiang↗
Multimodal Few-Shot Segmentation of Electron Micrographs
Scanning transmission electron microscopy (STEM) is one of the most used methods of analyzing the chemistry and composition of materials. By analyzing microstructures, these microscopes can help scientists better understand the molecular underpinnings of microelectronics, batteries, and more. However, STEM data can be difficult to interpret, so recent developments have been made in applications of machine learning to analyze these images. The PNNL-developed pyCHIP Classifier has achieved results in segmenting STEM these images via few-shot learning, a method which requires little data and human input, perfect for quickly analysis. In my internship I (Eli Meyers) investigated a multimodal improvement of this classifier by incorporating energy dispersive x-ray spectroscopy (EDS) data into the classification process for a more accurate segmentation. Furthermore, I encoded the spectral data by training a mass spectrometry encoder on the EDS data to extract a more meaningful representation of the data.
36 MATERIALS SCIENCE↗
Learning Knowledge Graph Hierarchies for Improving Few-Shot Classification
Explore the source record for details and available documents.
Computer science↗
On the effectiveness of neural operators at zero-shot weather downscaling
Machine-learning (ML) methods have shown great potential for weather downscaling. These data-driven approaches provide a more efficient alternative for producing high-resolution weather datasets and forecasts compared to physics-based numerical simulations. Neural operators, which learn solution operators for a family of partial differential equations, have shown great success in scientific ML applications involving physics-driven datasets. Neural operators are grid-resolution-invariant and are often evaluated on higher grid resolutions than they are trained on, i.e., zero-shot super-resolution. Given their promising zero-shot super-resolution performance on dynamical systems emulation, we present a critical investigation of their zero-shot weather downscaling capabilities, which is when models are tasked with producing high-resolution outputs using higher upsampling factors than are seen during training. To this end, we create two realistic downscaling experiments with challenging upsampling factors (e.g., 8x and 15x) across data from different simulations: the European Centre for Medium-Range Weather Forecasts Reanalysis version 5 (ERA5) and the Wind Integration National Dataset Toolkit. While neural operator-based downscaling models perform better than interpolation and a simple convolutional baseline, we show the surprising performance of an approach that combines a powerful transformer-based model with parameter-free interpolation at zero-shot weather downscaling. We find that this Swin-Transformer-based approach mostly outperforms models with neural operator layers in terms of average error metrics, whereas an Enhanced Super-Resolution Generative Adversarial Network-based approach is better than most models in terms of capturing the physics of the ground truth data. We suggest their use in future work as strong baselines.
17 WIND ENERGY↗
Naturally-meaningful and efficient descriptors: machine learning of material properties based on robust one-shot ab initio descriptors
Establishing a data-driven pipeline for the discovery of novel materials requires the engineering of material features that can be feasibly calculated and can be applied to predict a material’s target properties. Here we propose a new class of descriptors for describing crystal structures, which we term Robust One-Shot Ab initio (ROSA) descriptors. ROSA is computationally cheap and is shown to accurately predict a range of material properties. These simple and intuitive class of descriptors are generated from the energetics of a material at a low level of theory using an incomplete ab initio calculation. We demonstrate how the incorporation of ROSA descriptors in ML-based property prediction leads to accurate predictions over a wide range of crystals, amorphized crystals, metal–organic frameworks and molecules. We believe that the low computational cost and ease of use of these descriptors will significantly improve ML-based predictions.
36 MATERIALS SCIENCE↗
Power Grid Contingency Analysis with Machine Learning: A Brief Survey and Prospects
We briefly review previous applications of machine learning (ML) in power grid analyses and introduce our ongoing effort toward developing a generative-adversarial (GA) model for fast and reliable grid contingency analyses. According to our review, the persisting limitation of traditional ML techniques in grid analyses is the need for an exhaustive amount of training data for model generalization and accurate predictions. GA models overcome this limitation by first learning true data distribution from a small training set, from which new samples assimilating true data are generated with some variations. Subsequently, GA models can transfer learn or super-generalize with increased accuracy, that is, accurately predict n - (k + 2) contingencies from a small n - k training set and generated n - (k + 1) data. The joint effort between Idaho National Lab and Florida State University strives to develop a zero-shot and deep learning-based contingency analysis tool, named Smart Contingency Analysis Neural Network (SCANN), by leveraging the aforementioned advantages of GA models. The basic architecture of SCANN stems from the Latent Encoding of Atypical Perturbations network combined with an adversarial network, and it is designed to generate imbalanced power flow data from learned true data distributions for prediction purposes. Here we also introduce the abstract concept of resilience-chaos plots, a new resilience characterization tool proposed to complement SCANN by aiding in the assessment of large amounts of high-order contingency predictions.
24 POWER TRANSMISSION AND DISTRIBUTION↗
Dynamic in-context learning with conversational models for data extraction and materials property prediction
The advent of natural language processing and large language models (LLMs) has revolutionized the extraction of data from unstructured scholarly papers. However, ensuring data trustworthiness remains a significant challenge. In this paper, we introduce PropertyExtractor, an open-source tool that leverages advanced conversational LLMs such as Google gemini-pro and OpenAI gpt-4, blends zero-shot with few-shot in-context learning, and employs engineered prompts for the dynamic refinement of structured information hierarchies—enabling autonomous, efficient, scalable, and accurate identification, extraction, and verification of material property data. Our tests on material data demonstrate precision and recall that exceed 95% with an error rate of ∼9%, highlighting the effectiveness and versatility of the toolkit. Finally, databases for 2D material thicknesses, a critical parameter for device integration, and energy bandgap values are developed using PropertyExtractor. In particular, for the thickness database, the rapid evolution of the field has outpaced both experimental measurements and computational methods, creating a significant data gap. Our work addresses this gap and showcases the potential of PropertyExtractor as a reliable and efficient tool for the autonomous generation of various material property databases, advancing the field.
Ekuma, Chinedu E. (ORCID:0000000258527556)↗
EdgeAI: Machine learning via direct attached accelerator for streaming data processing at high shot rate x-ray free-electron lasers
We present a case for low batch-size inference with the potential for adaptive training of a lean encoder model. We do so in the context of a paradigmatic example of machine learning as applied in data acquisition at high data velocity scientific user facilities such as the Linac Coherent Light Source-II x-ray Free-Electron Laser. We discuss how a low-latency inference model operating at the data acquisition edge can capitalize on the naturally stochastic nature of such sources. We simulate the method of attosecond angular streaking to produce representative results whereby simulated input data reproduce high-resolution ground truth probability distributions. By minimizing the mean-squared error between the decoded output of the latent representation and the ground truth distributions, we ensure that the encoding layers and resulting latent representation maintains full fidelity for any downstream task, be it classification or regression. We present throughput results for data-parallel inference of various batch sizes, some with throughput exceeding 100 k images per second. We also show in situ training below 10 s per epoch for the full encoder–decoder model as would be relevant for streaming and adaptive real-time data production at our nation’s scientific light sources.
97 MATHEMATICS AND COMPUTING↗
BuildingsBench: A Benchmark for Universal Building Load Forecasting [SWR-23-51]
The residential and commercial building stock in the United States is responsible for a significant percentage of energy consumption and greenhouse gas emissions. Electrification of end-uses, as well as decarbonizing the electrical grid through renewable energy sources such as solar and wind, constitutes the pathway to zero-emission buildings. Forecasting day-ahead building energy consumption is an integral part of this solution. Currently, specialized forecasting models are hand-made for each individual building, which is time-consuming, expensive, and leads to duplicated efforts. BuildingsBench is a Python software framework for training and comparing generalized machine learning models for universal building load forecasting. This challenge tasks a single foundational model to generalize its forecasts for a wide variety of buildings, across geographic regions, building types, weather patterns, and more. This software provide code for pre-training such models and subsequently evaluating their performance on a suite of hundreds of diverse real and synthetic buildings. BuildingsBench is a platform for: - Large-scale pretraining with the synthetic Buildings-900K dataset for short-term load forecasting (STLF). Buildings-900K is statistically representative of the entire U.S. building stock and is extracted from the NREL End-Use Load Profiles database. - Benchmarking on two tasks evaluating generalization: zero-shot STLF and transfer learning for STLF. We provide an index-based PyTorch Dataset for large-scale pretraining, easy data loading for multiple real building energy consumption datasets as PyTorch Tensors or Pandas DataFrames, simple (persistence) to advanced (transformer) baselines, metrics management, and more.
Emami, Patrick↗
DefectTrack: a deep learning-based multi-object tracking algorithm for quantitative defect analysis of in-situ TEM videos in real-time
Abstract In-situ irradiation transmission electron microscopy (TEM) offers unique insights into the millisecond-timescale post-cascade process, such as the lifetime and thermal stability of defect clusters, vital to the mechanistic understanding of irradiation damage in nuclear materials. Converting in-situ irradiation TEM video data into meaningful information on defect cluster dynamic properties (e.g., lifetime) has become the major technical bottleneck. Here, we present a solution called the DefectTrack , the first dedicated deep learning-based one-shot multi-object tracking (MOT) model capable of tracking cascade-induced defect clusters in in-situ TEM videos in real-time. DefectTrack has achieved a Multi-Object Tracking Accuracy (MOTA) of 66.43% and a Mostly Tracked (MT) of 67.81% on the test set, which are comparable to state-of-the-art MOT algorithms. We discuss the MOT framework, model selection, training, and evaluation strategies for in-situ TEM applications. Further, we compare the DefectTrack with four human experts in quantifying defect cluster lifetime distributions using statistical tests and discuss the relationship between the material science domain metrics and MOT metrics. Our statistical evaluations on the defect lifetime distribution suggest that the DefectTrack outperforms human experts in accuracy and speed.
42 ENGINEERING↗
Integrated Design of a Robust System for Inertial Fusion Energy
For inertial fusion energy (IFE) to be successful, designs need to work robustly including both target physics and engineering considerations. For a power plant to be reliable and have a high operating capacity factor, the integrated design of the target, driver, target fabrication, target injection system, and chamber all need to work together in way that is robust and repeatable to expected variations. One of the lesson’s learned from our experience at NIF is that designs that are sensitive have not performed as expected – even in single shot mode. The lesson’s learned from NIF should be folded into creating integrated designs for IFE and for evaluating the tradeoffs between the different parts of the IFE system. To do so, we propose using the NIF 1.3 MJ yield shot/design (N210808) as a starting point of a study of the feasibility of indirect drive designs for IFE. If the hohlraum inefficiency precludes indirect drive, the study would include direct drive or fast ignition versions of this design.
70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
Knowledge Oriented Graph Unified Transformer (KOGUT) v0.1
KOGUT — Knowledge Oriented Graph Unified Transformer KOGUT implements the Relational Graph Transformer (RelGT) architecture for knowledge graph link prediction in biological domains, with a primary focus on microbial growth media prediction. While the original RelGT (arXiv:2505.10960) targets relational tables, time series, and multi-table databases, KOGUT adapts this architecture for heterogeneous biological knowledge graphs, providing first-in-class AI predictive models for microbial cultivation. Key Adaptations Beyond Original RelGT: - Knowledge Graph Focus: Applied to biological KGs with semantic node types (taxa, chemicals, media, phenotypes, environments) versus generic relational database tables, trained on the KG-Microbe knowledge graph (1.3M entities, 2.9M edges, 24 relation types). - Multimodal Node Encoding: Integrates node labels, categories, descriptions, and synonyms from KG metadata through learned embedding layers—adapting relational column features to graph node attributes with textual semantics. - Extended K-Hop Subgraph Strategy: Optimized neighborhood sampling (3-hop default, configurable up to 200 nodes) tuned for sparse biological networks, building on the original local-global attention framework with biological relation preservation. - Biolink Predicate Preservation: Type-specific transformations for 24 biological edge semantics (occurs_in, consumes, produces, has_phenotype, subclass_of) beyond standard relational foreign keys, enabling multi-relation link prediction. - Inductive Learning Support: Enables zero-shot predictions for novel taxa through feature-based embeddings (temperature, oxygen requirements, gram stain, cell shape), extending the original transductive relational benchmark scope to uncultured microorganisms. CheapSOTA Performance Optimizations (This Distribution): - VQ-EMA Centroid Attention: Vector quantization with exponential moving average for improved global context modeling (+5-10% MRR improvement). - HDF5 Precomputed Data Loading: One-time preprocessing of k-hop subgraphs to eliminate redundant graph traversals (2-5× training speedup). - Distributed Data Parallel Training: Multi-GPU support for scaling to larger knowledge graphs (tested on 4× NVIDIA A100 GPUs at NERSC Perlmutter). - Mixed Precision Training: Automatic mixed precision (AMP) for memory efficiency and faster training. Advantages Over Standard Knowledge Graph Embedding Models: Combines RelGT's proven multi-element tokenization (features, type, hop, structure) with graph-native biological representations, enabling interpretable link prediction across heterogeneous entities that standard embedding models (TransE, RotatE, ComplEx) and table-based transformers cannot directly model. Achieves near-perfect performance on microbial growth media prediction (MRR: 0.9966, Precision@1: 0.9932, Hit@10: 1.0000) while maintaining explainability through attention-based reasoning over biological pathways. Training Data: - KG-Microbe merged knowledge graph: 1,379,337 nodes, 2,960,472 edges - 24 biological relation types including taxonomic hierarchies, metabolic interactions, phenotype associations, and environmental relationships - Primary prediction task: Growth media suitability for microbial taxa (biolink:occurs_in, 50K edges) - Multi-relation capability: Predicts links for any of the 24 relation types, including chemical consumption/production, phenotype associations, and taxonomic classification Citation: Original RelGT Architecture: Dwivedi et al., "Relational Graph Transformer", arXiv:2505.10960, 2025 KOGUT Implementation: Knowledge Oriented Graph Unified Transformer for Microbial Growth Media Prediction Developed at Lawrence Berkeley National Laboratory (LBNL) Trained on NERSC Perlmutter supercomputer
Joachimiak, Marcin [Lawrence Berkeley National Lab↗
Investigating performance and variability of NIF ICF experiments with deep learning
The parameter space involved in designing an inertial confinement fusion shot at the National Ignition Facility (NIF) is massively multi-dimensional and the cost of a single shot makes a comprehensive set of sensitivity studies in the laboratory impractical. The use of machine learning to overcome these challenges has gained popularity and has had several successful applications by the scientific community. We extend on these efforts by training a neural network (NN) on information about the experimental design, engineering elements, and drive asymmetry to predict with uncertainty the neutron yield of an experiment. We find the measured and model predicted values are in good agreement, with an R 2 value of 0.91 for a randomly selected test dataset. Almost all the predicted 95% credible intervals contain the corresponding measured value for both training and test datasets. We identify correlations picked up by the NN between the shot design, yield, and variability and use them to motivate shot sensitivity studies. The first shot to exceed the Lawson-like ignition criteria (N210808) was conducted at the NIF and subsequent shots studied the design’s robustness. In a follow-up shot to N210808, our model predicts capsule quality to be the main performance degradation mechanism that prevented the shot from repeating previous performance levels. Shot N221204 was the first shot to exceed a target energy gain of 1. Our model predicts increased yield with reduced coast time for a N221204 study and greater variability for designs with lower peak powers at constant yield. The model’s fast prediction speed and uncertainty prediction are useful for identifying interesting design paths that could warrant further investigation with conventional simulations to search for robust high yield designs.
70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
Automated experimental design of safe rampdowns via probabilistic machine learning
Abstract Typically the rampdown phase of a shot consists of a decrease in current and injected power and optionally a change in shape, but there is considerable flexibility in the rate, sequencing, and duration of these changes. On the next generation of tokamaks it is essential that this is done safely as the device could be damaged by the stored thermal and electromagnetic energy present in the plasma. This works presents a procedure for automatically choosing experimental rampdown designs to rapidly converge to an effective rampdown trajectory. This procedure uses probabilistic machine learning methods paired with acquisition functions taken from Bayesian optimization. In a set of 2022 experiments at DIII-D, the rampdown designs produced by our method maintained plasma control down to substantially lower current and energy levels than are typically observed. The actions predicted by the model significantly improved as the model was able to explore over the course of the experimental campaign.
70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
femto-PIXAR: a self-supervised neural network method for reconstructing femtosecond X-ray free electron laser pulses
X-ray Free Electron Lasers (X-FELs) operate in a wide range of lasing configurations for a broad variety of scientific applications at ultrafast time-scales such as structural biology, materials science, and atomic and molecular physics. Shot-by-shot characterization of the X-FEL pulses is crucial for analysis of many experiments as well as tuning the X-FEL performance. However, for the weak pulses found in advanced configurations, e.g. those needed for coherent, two-pulse studies of quantum materials, there is no current method for reliably resolving pulse profiles. Here we show that a physics-based U-net model can reconstruct the individual pulse power profiles for sub-picosecond pulse separation without the need for simulations. Using experimental data from weak X-FEL pulse pairs, we demonstrate we can learn the pulse characteristics on a shot-by-shot basis when conventional methods fail.
43 PARTICLE ACCELERATORS↗
Coordinate-Based Seismic Interpolation in Irregular Land Survey: A Deep Internal Learning Approach
Physical and budget constraints often result in irregular sampling, which complicates accurate subsurface imaging. Preprocessing approaches, such as missing trace or shot interpolation, are typically employed to enhance seismic data in such cases. Recently, deep learning has been used to address the trace interpolation problem at the expense of large amounts of training data to adequately represent typical seismic events. Nonetheless, most research in this area has focused on trace reconstruction, with little attention having been devoted to shot interpolation. Furthermore, existing methods assume regularly spaced receivers/sources failing in approximating seismic data from real (irregular) surveys. This work presents a novel shot gather interpolation approach which uses a continuous coordinate-based representation of the acquired seismic wavefield parameterized by a neural network. The proposed unsupervised approach, which we call coordinate-based seismic interpolation (CoBSI), enables the prediction of specific seismic characteristics in irregular land surveys without using external data during neural network training. Importantly, experimental results on real and synthetic 3-D data validate the ability of the proposed method to estimate continuous smooth seismic events in the time-space and frequency-wavenumber domains, improving sparsity or low-rank-based interpolation methods.
58 GEOSCIENCES↗