8:00-9:20am: Arrival + Breakfast
9:20-9:50am: Welcome and Introduction
9:50-12:00pm: Session 1: Do ML models benefit from architectures informed by domain knowledge?
9:50-10:35: Tess Smidt (Invited Talk)
10:35-10:55: Julie Zhu (Conformal Predictions under Lévy-Prokhorov Distribution Shifts: Robustness to Local and Global Perturbations)
10:55-11:15: Yo Akiyama (Scaling Down and Expanding the Scope of Protein Language Modeling with MSA Pairformer)
11:15-12:00: Sergey Ovchinnikov (Invited Talk)
12:00-1:30pm: Lunch Break
1:30-3:40pm: Session 2: Can ML generate interpretable, novel hypotheses that align with or challenge existing theories?
1:30-2:15: Fabian Ruehle (Invited Talk)
2:15-2:35: Maria Carreira (Leveraging Generative Modeling to Analyze Multiple Related Cryo-EM Datasets)
2:35-2:55: Berthy Feng (Computational Imaging for Astronomy Using Data-Driven and Physics-Based Priors)
2:55-3:40: Sendhil Mullainathan (Invited Talk)
3:40-4:00pm: Coffee Break
4:00-6:00pm: Reception and discussions
6:00-7:30pm: Dinner
7:30pm-: Social team building activity
7:30-8:30am: Breakfast
8:30-9:30am: Lightning Talks+Forum
Teddy Koker, Julia Balla, Kevin Yu, Spencer Wyant, Miguel Steiner
9:30-12:00pm: Session 3: What kinds of scientific problems can off-the-shelf ML models learn effectively, and where do they fall short?
9:30-10:15: Cuong Nguyen (Invited Talk)
10:15-10:35: Kate Fisher (Towards a Bridge between Theory and Practice: Sobolev Training in ML applied to Quantum Chemistry)
10:35-10:55: Sauradeep Majumdar (ML-driven Discovery of Potential Nanoporous Materials for Energy-related Applications)
10:55-11:15: Jeff Brown (Evaluating LLMs for Wet-lab Experimental Planning and Feasibility Assessment)
11:15-12:00: Michael Brenner (Invited Talk)
12:00-1:30pm: Lunch Break
1:30-3:35pm: Session 4: How can we test whether ML models reflect known scientific principles?
1:30-2:15: Nima Dehmamy (Invited Talk)
2:15-2:35: Jonas Rein (Discovery of Missing Links in Chemical Synthesis)
2:35-2:55: Jingyi Qiu (Measuring Rhetorical Style in Scientific Writing with LLM Personas)
2:55-3:15: Nate Somers (Engineering Conformational Ensembles to Control Molecular Properties)
3:15-3:35: Zongyi Li (Neural Operator for Scientific Computing)
3:35-4:15pm: Coffee Break
4:15-5:00: Wrap up and takeaways
5:00: Departure