Description: Foundation models, most prominently large language models (LLMs), are rapidly revolutionizing research tasks, from ideation and writing, to coding and data analysis. Astronomy-specific foundation models, trained on carefully curated, massive archival datasets, are providing new approaches for wrangling, exploring, and analyzing current and next-generation surveys. Meanwhile, AI agents, built on top of these models, are achieving previously unimaginable gains in autonomy and complex reasoning. This rapidly evolving AI landscape offers enormous potential to accelerate astronomy research tasks and may ultimately create a new paradigm for how astronomers interact with data and make discoveries.
Location: Salt Lake City at the Salt Palace Convention Center
Speakers
This special session will:
Review advances and applications of foundation models in astronomy research;
Highlight recently developed public models and benchmarks;
Discuss opportunities and pitfalls of agentic AI use in astronomy research workflows;
Solicit community feedback on foundation model adoption: outstanding challenges, goals, and needs.
Format:
50 minutes of talks highlighting foundation model development and applications (4 invited speakers + 2 5 minute invited talks)
20-minute panel on the opportunities and pitfalls of agentic AI in research
10-minute moderated audience Q&A and discussion
10-minute 1-minute Poster presentations.