Talks

• 2026-09-04 LLM Agents for Science: Vibe Coding, Autonomous Scientists, and BeyondUpcoming Fast Machine Learning for Science Conference 2026 · QI Auditorium, UC San Diego · Birds-of-a-Feather session, with Eric Moreno (MIT) and Nan Huang (UC San Diego) An open discussion on LLM agents in high-energy-physics workflows: where vibe coding breaks down, how far autonomous scientific agents actually are, the risks around hallucinated results and reproducibility, and what benchmarks would be needed to trust agent-generated science. Programme ↗
• 2026-09-02 SIDERIUS: Resource-Aware Multi-Agent System for Automated Machine LearningUpcoming Fast Machine Learning for Science Conference 2026 · QI Auditorium, UC San Diego · Invited talk A resource-aware multi-agent system for automated machine learning — coordinating scientific reasoning, model proposal, implementation, validation, tuning and interpretation under limited compute — and the lessons that motivated SciTra. Programme ↗
• 2026-06-21 Lecture 4: LLM, AI Agent, and Vibe Coding 0νββ AI Summer School · ISEB Colloquium, UC Irvine · Invited lecture A hands-on introduction to large language models, agents and vibe coding for physicists, given at the Neutrino 2026 satellite school on AI for neutrinoless double-beta-decay experiments. Programme ↗Slides ↗
• 2026-03-16 TIDMAD: A Sandbox for Physics-Informed LLM Agents in Dark Matter Searches APS Global Physics Summit 2026 · Session: AI and ML for Materials Characterization and Spectroscopies · Invited talk On TIDMAD, the open ultra-long time-series benchmark from the ABRACADABRA axion experiment, and first steps toward physics-informed LLM agents that explore denoising strategies and characterise the structure and origin of detector noise. With Aobo Li. Abstract ↗Slides ↗
• 2025-03-18 Measuring the Solar pp Neutrino Flux using Electron Recoil Data from XENONnT APS Global Physics Summit 2025 · Session: Solar, Supernova, and Galactic Neutrinos · Contributed talk On constraining the dominant low-energy electron-recoil background in XENONnT through dedicated radon calibrations and spatially optimised data selection, and the resulting sensitivity to the solar pp neutrino flux. Abstract ↗
• 2023-04-15 Status and Prospects of the Electron Recoil Channel in XENONnT 2023 APS April Meeting · Session: Dark Matter III · Contributed talk XENONnT's first low-energy electron-recoil search: 1.16 tonne-years of exposure, no excess above background, and new limits on solar axions, an enhanced neutrino magnetic moment, and bosonic dark matter. Presenting Phys. Rev. Lett. 129, 161805. Abstract ↗Paper ↗

Slides for the September talks will be posted after the conference.

• 2026-03-18 Machine Learning Methods at the Intersection of Nuclear and Neutrino Physics APS Global Physics Summit 2026 · Contributed session (APR-R87) · Chair Sponsored by the APS Topical Group on Data Science (GDS) and the Division of Nuclear Physics (DNP). Session ↗