Program Agenda
July 31 – August 1, 2026 • Harvard University
Day 1 — Friday, July 31, 2026
Short Courses, Data Challenge Demo, & Evening Poster Session and Reception
Three morning short courses (8:00 AM – 12:00 PM) and three afternoon short courses (1:00 PM – 5:00 PM) run in parallel. Participants may register for one morning and one afternoon course. The day closes with a demo session of the challenge awardees and a welcome reception in the evening.
Venues — Maxwell-Dworkin (MD): 33 Oxford Street • Divinity School (DS): 45 Francis Ave, Cambridge, MA 02138.
| Time | Program Item | Title | Instructor(s) |
|---|---|---|---|
| 8:00 – 8:30 AM | Registration & Breakfast📍 MD — G100 Lobby | — | — |
| 8:00 AM – 12:00 PM | SC1 — Morning Track A📍 MD — Room G-115 | Agentic AI: From Zero to Infinity | Tian Zheng (Columbia) |
| 8:00 AM – 12:00 PM | SC2 — Morning Track B📍 MD — Room G-135 | Theory of Scaling in Modern Deep Learning | Soufiane Hayou (JHU); Nikhil Ghosh (Flatiron) |
| 8:00 AM – 12:00 PM | SC3 — Morning Track C📍 MD — Room G-125 | An Overview of LLMs for Statisticians | Linjun Zhang (Rutgers) |
| 10:00 – 10:15 AM | Coffee Break📍 MD — G100 Lobby | — | — |
| 12:00 – 1:00 PM | Lunch Break📍 MD — G100 Lobby | — | — |
| 1:00 – 5:00 PM | SC4 — Afternoon Track A📍 MD — Room G-135 | Reinforcement Learning: Foundations & Applications | Chengchun Shi (LSE) |
| 1:00 – 5:00 PM | SC5 — Afternoon Track B📍 MD — Room G-125 | AI in Accelerating Clinical Trials & Drug Development | Alexia Iasonos (MSKCC); John O'Quigley (UCL) |
| 1:00 – 5:00 PM | SC6 — Afternoon Track C📍 MD — Room G-115 | Theory for Diffusion Models: Continuous & Discrete | Sitan Chen (Harvard) |
| 3:00 – 3:15 PM | Coffee Break📍 MD — G100 Lobby | — | — |
| 5:15 – 5:45 PM | Data Challenge Demo📍 DS — James Room | Demo by the Challenge awardees | — |
| 5:45 – 7:45 PM | Welcome Reception📍 DS — Café Commons | Poster session, open to all attendees and instructors | — |
Day 2 — Saturday, August 1, 2026
Main Conference Day
Each of Sessions 1, 2, and 3 runs 1 hour 30 minutes and contains 5 talks: 2 invited talks (25 min each, 5 min Q&A) followed by 3 short paper award talks (8 min each, 2 min Q&A). The lightning talk session is 30 minutes by 9 stellar abstract award winners (3 min each).
Venues — Science Center (SC): 1 Oxford Street (main conference) • Northwest (NW): 52 Oxford Street (lunch), Cambridge, MA 02138.
| Time | Talk | Speaker |
|---|---|---|
| 8:00 – 8:55 AM | Breakfast📍 Science Center | — |
| 8:55 – 9:00 AM | Welcome and Opening Remarks📍 Science Center — Hall C | — |
| Session 1 · 9:00 – 10:30 AM · Chair: Wenyi Wang · 📍 Science Center — Hall C | ||
| 9:00 – 9:30 AM | Invited TalkTraining in Imagination or Learning by Telling the Truth — A Debate on World Model | Eric XingMBZUAI & Carnegie Mellon University |
| 9:30 – 10:00 AM | Invited TalkLabOS: The AI-XR Co-Scientist That Sees and Works With Humans | Mengdi WangPrinceton University |
| 10:00 – 10:10 AM | Paper Award TalkGenerative AI and Entrepreneurship: A Spatial Synthetic Difference-in-Differences Approach | Jeff CaiUniversity of Notre Dame |
| 10:10 – 10:20 AM | Paper Award TalkA Deep Generative Framework for Right-Censored Survival Data with Distribution-Matching Regularization | Yanran Li, Zexi CaiColumbia University |
| 10:20 – 10:30 AM | Paper Award TalkCrossHONA: Cross-species HOmologous and Non-homologous gene-aware framework for transcriptomics integration and Annotation | Ruohan WangCenter for Computational Molecular Biology, Brown University |
| 10:30 – 11:00 AM | Coffee Break📍 Science Center | — |
| Session 2 · 11:00 AM – 12:30 PM · Chair: Linjun Zhang · 📍 Science Center — Hall C | ||
| 11:00 – 11:30 AM | Invited TalkFrom Entropy to Epiplexity: Rethinking Information for Computationally Bounded Intelligence | Andrew WilsonNew York University |
| 11:30 AM – 12:00 PM | Invited TalkSynthetic Data for Black-Box Validation | Martin WainwrightMassachusetts Institute of Technology |
| 12:00 – 12:10 PM | Paper Award TalkSTAVAT: Stratified Anytime-Valid Atom Sampling for LLM Tier Placement | Patrick VosslerUniversity of California, San Francisco |
| 12:10 – 12:20 PM | Paper Award TalkVariance-aware Reward Modeling with Anchor Guidance | Shuxing FangThe Hong Kong Polytechnic University |
| 12:20 – 12:30 PM | Paper Award TalkMembership Inference via Pairwise Likelihood Ratios | Shengjie NiuThe Hong Kong Polytechnic University |
| 12:30 – 1:40 PM | Lunch Break & Round Tables📍 Northwest Building — Basement B100 (~6 min walk from Science Center) | — |
| 1:40 – 2:00 PM | Award Ceremony📍 Science Center — Hall C | — |
| Session 3 · 2:00 – 3:30 PM · Chair: Edgar Dobriban · 📍 Science Center — Hall C | ||
| 2:00 – 2:30 PM | Invited TalkFeature Learning and the Linear Representation Hypothesis for Steering and Monitoring LLMs | Misha BelkinUniversity of California San Diego |
| 2:30 – 3:00 PM | Invited TalkStatlib: A Statistical Library for Research, Education, and Beyond in an AI World | Rajarshi MukherjeeHarvard T.H. Chan School of Public Health |
| 3:00 – 3:10 PM | Paper Award TalkWahkon: A Statistically Principled Deep RKHS Superposition Network | Yongkai ChenHarvard University |
| 3:10 – 3:20 PM | Paper Award TalkAI4BayesCode: From Natural Language Descriptions to Validated Modular Stateful Bayesian Samplers | Jungang ZouDepartment of Biostatistics, Columbia University |
| 3:20 – 3:30 PM | Paper Award TalkDensity-Informed Pseudo-Counts for Calibrated Evidential Deep Learning | Pietro CarlottiUniversity of Texas at Austin |
| 3:30 – 4:00 PM | Coffee Break📍 Science Center | — |
| 4:00 – 5:00 PM | Panel Discussion📍 Science Center — Hall C | Moderated panel + audience Q&A |
| 5:00 – 5:30 PM | Stellar Abstract Lightning Talks📍 Science Center — Hall C | 9 × 3-min Stellar Abstract Award talks |
| 5:30 – 5:35 PM | Closing Remarks📍 Science Center — Hall C | — |
| 5:35 – 7:30 PM | Poster Session & Reception Supported by Jump Trading📍 Science Center | Posters, food & beverages |