29 October 2024 in APL Machine Learning
The specialized language and complex concepts in physics pose significant challenges for information extraction through Natural Language Processing (NLP). Central to effective NLP applications is the text embedding model, which converts text into dense vector representations for efficient information retrieval and semantic analysis. In this work, we introduce PhysBERT, the first physics-specific text embedding model. Pre-trained on a curated corpus of 1.2 × 106 arXiv physics papers and fine-tuned with supervised data, PhysBERT outperforms leading general-purpose models on physics-specific tasks, including the effectiveness in fine-tuning for specific physics subdomains.
15 November 2022 in Physical Review Accelerators and Beams
The ALS-U is the upgrade of the existing Lawrence Berkeley National Laboratory Advanced Light Source to a diffraction-limited soft x-ray light source. Here we present the lattice correction studies and commissioning simulations demonstrating that the proposed machine design can be expected to deliver the intended performance when realistic errors and perturbations are fully accounted for. Critical to this demonstration are the high-fidelity, realistic simulations of the beam-based alignment process (both in turn-by-turn mode during early commissioning and with stored beam) that are now made possible by the Toolkit for Simulated Commissioning. In addition to presenting a statistical performance analysis based on a large number of lattice error realizations, we also study the range of further improvements that can be obtained by fine-tuning the correction chain to individual error seeds, mimicking the approach one would follow once the machine is built.
30 July 2024 in Physical Review Accelerators and Beams
Past research at the Advanced Light Source (ALS) provided a proof-of-principle demonstration that deep learning methods could be effectively employed to compensate for the significant perturbations to the transverse electron beam size induced by user-controlled adjustments of the insertion devices. However, incorporating these methods into the ALS’ daily operations has faced notable challenges. The complexity of the system’s operational requirements and the significant upkeep demands has restricted their sustained application during user operation. Here, we introduce the development of a more robust neural network (NN)-based algorithm that utilizes a novel online fine-tuning approach and its systematic integration into the day-to-day machine operations. Our analysis emphasizes the process of NN model selection, demonstrates the superior performance of the NN-based method over traditional feedback methods, and examines the effectiveness and resilience of the new algorithm during user-operation scenarios. Published by the American Physical Society 2024
10 October 2019 in Physical Review Accelerators and Beams
We present a new accelerator toolbox (AT)-based toolkit for simulating the commissioning of light-source storage rings. The toolkit provides a framework for supporting high-level scripts to represent with realism the various procedures (e.g., orbit and optics correction, beam-based alignment, etc.) encountered during commissioning and is designed to mirror as closely as possible the reality as seen from the control room. Emphasis is placed on the inclusion of a comprehensive set of error sources and faithful modeling of beam diagnostics. The toolkit capabilities are demonstrated in an application to the recent design and commissioning studies of the Advanced Light Source Upgrade (ALS-U) Accumulator Ring, a short-time successful commissioning of which will be critical to the overall ALS-U project success.
4 February 2026 in APL Machine Learning
Operating large-scale scientific facilities requires coordinating diverse subsystems, translating operator intent into precise hardware actions, and maintaining strict safety oversight. Language model-driven agents offer a natural interface for these tasks, but most existing approaches are not yet reliable or safe enough for production use. In this paper, we introduce Osprey, a framework for using agentic AI in large, safety-critical facility operations. Osprey is built around the needs of control rooms and addresses these challenges in four ways. First, it uses a plan-first orchestrator that generates complete execution plans, including all dependencies, for human review before any hardware is touched. Second, a coordination layer manages complex data flows, keeps data types consistent, and automatically downsamples large datasets when needed. Third, a classifier dynamically selects only the tools required for a given task, keeping prompts compact as facilities add capabilities. Fourth, connector abstractions and deployment patterns work across different control systems and are ready for day-to-day use. We demonstrate the framework through two case studies: a control-assistant tutorial showing semantic channel mapping and historical data integration and a production deployment at the Advanced Light Source, where Osprey manages real-time operations across hundreds of thousands of control channels. These results establish Osprey as a production-ready framework for deploying agentic AI in complex, safety-critical environments.
14 April 2025 in Physical Review Accelerators and Beams
Accelerator physics presents unique challenges for natural language processing (NLP) due to its specialized terminology and complex concepts. A key component in overcoming these challenges is the development of robust text embedding models that transform textual data into dense vector representations, facilitating efficient information retrieval and semantic understanding. In this work, we introduce AccPhysBERT, a sentence embedding model fine-tuned specifically for accelerator physics. Our model demonstrates superior performance across a range of downstream NLP tasks, surpassing existing models in capturing the domain-specific nuances of the field. We further showcase its practical applications, including semantic paper-reviewer matching and integration into retrieval-augmented generation systems, highlighting its potential to enhance information retrieval and knowledge discovery in accelerator physics.