Seminars

Liver toxicity can emerge late in drug development when early assays fail to capture human biology, exposure or injury mechanisms. This webinar introduces Axiom’s hepatic profiling approach, combining multicellular liver systems, high-content imaging, transcriptomics, proteomics and ADME measurements. Its reasoning agent integrates those results with partner assays, chemical-series context, exposure estimates and clinical evidence to assess mechanistic liver-safety risks. Rather than using a single cell-toxicity score, the approach develops interpretable, series-specific structure–toxicity hypotheses, compares analogues, ranks compounds and suggests follow-up experiments or structural changes. Performance on established liver-injury reference compounds and newer molecules is examined through case studies, showing how multimodal evidence can inform medicinal-chemistry decisions and predicted therapeutic index.

Ilaria Ferlenghi discusses a career spanning structural biology, cryo-electron microscopy, vaccine research and development, and the growing role of artificial intelligence and machine learning in scientific innovation and entrepreneurship.

Deep networks are often thought of as black boxes. Their ability to encompass vast swathes of knowledge indeed makes them hard to explain. Yet many of their behaviours — generalization under overparametrization, grokking, OOD failures, neural scaling laws — recur across architectures and scales, and each, however surprising, can be reproduced and explained in controlled settings. I will review these efforts to identify and explain the universal phenomena of deep learning, and suggest that an LLM may amount to a sum of such tractable sub-phenomena, interpolative in nature. Finally, leaving scientific rigor aside, I'll argue that what separates this prosaic picture from the apparent magic of LLMs may well be the industrial scale of compute and human labour behind it, and that AGI in its deeper extrapolative sense may be much further away than claimed.

Jie Yang presents BRIDGE, a multilingual evaluation benchmark built from real clinical tasks and more than one million samples derived from electronic health records. The benchmark addresses the gap between simplified examination-style tests and the complexity of clinical data, while enabling comparison of rapidly changing language models. Evaluation of 95 models involved over 24,000 experiments and 39 million predictions. Results vary substantially by model family, task and language, with some open-source systems matching proprietary models. The study also finds that chain-of-thought prompting frequently reduces accuracy on these tasks and examines stigmatizing language produced during model reasoning at large scale.

Matthias Vigl examines whether the computational scaling strategies behind modern machine learning can benefit particle physics. Empirical scaling laws relate model performance to computing resources and help balance model size against training-data volume. For a fixed dataset, increasing model capacity beyond the interpolation threshold can improve generalization through double descent, but gains eventually encounter limits imposed by the data. The talk then considers scaling data and model size together, deriving compute-optimal relations for transformer-based jet taggers trained on as many as billions of simulated jets. It studies how training hyperparameters and input representations change these relations. The observed behaviour suggests that larger computing budgets could yield useful gains in particle physics, especially with a move toward more general-purpose foundation models.

Jan Kosinski explains how AlphaFold-style methods predict protein complexes and why their reliance on evolutionary information makes interactions between host and pathogen proteins especially difficult. The talk reviews large-scale studies, interpreting both successful predictions and failures rather than treating reported success rates as universal performance. It then examines ways to improve predictions through broader sampling, altered sequence alignments and experimental restraints, including crosslinking mass spectrometry, with influenza A virus as a worked research example. The session is suitable for students and researchers interested in host-pathogen interactions and structural bioinformatics; basic knowledge of protein structures and sequence alignments is useful, but prior AlphaFold experience is unnecessary.

Robot missions require both collision-free trajectories and logical task ordering, such as collecting a resource before entering a restricted region. Augmented graphs of convex sets combine continuous trajectory optimization and discrete task sequencing in one problem. A layered graph constructed from an exact convex partition of free space selects an optimal task order and trajectory through a shortest-path calculation, exact up to finite Bézier parameterization. Its structure corresponds to Bellman–Held–Karp dynamic programming for the travelling-salesperson problem, retaining singly exponential worst-case complexity and improving exponentially on general temporal-logic tools. A library of specifications covers ordered collection, alternative keys, combined prerequisites, timing and conditional logic with correctness guarantees. Numerical benchmarks demonstrate exponential speedups and near-global optimality. Possible extensions include moving environments, safety-aware planning with conformal prediction sets and a large galactic-survey benchmark.

Yixuan He (Arizona State University) considers graph neural networks for social, financial and biological networks in which edge signs and directions contain information needed for clustering and prediction. After an introduction to graph neural networks, the talk explains how signed and directed methods preserve relationships lost when every connection is treated as unsigned and undirected. The seminar then asks when a representation learned before the final task can transfer to a different graph or to a setting with few labels. Two theoretical perspectives address transfer across graph domains and the use of link prediction for linear community detection. TopoDIG and TopoSIGN combine encoders that preserve edge structure with persistent topological features, enabling pre-training and prompt-based adaptation on directed and signed networks.

Srinivasan Keshav (University of Cambridge) presents an interface designed to help ecologists use AI for remote sensing despite unfamiliarity with embeddings and algorithms. The Tessera Embeddings Explorer lets practitioners privately explore a region of interest, create habitat maps and perform regressions using their domain expertise and ground-truth knowledge. The talk demonstrates the tool and explains the design principles and the software architecture that follows from them, connecting usable ecological workflows with AI representations. Keshav is the Robert Sansom Professor of Computer Science at Cambridge.

Alok Puranik (Jane Street) examines data mixing for large language models: allocating model capacity and computation among training sources. A small model’s preferred data mix may differ from that of a larger model, yet the cost of large-model training forces many decisions to depend on smaller experiments. Scaling laws can predict performance as size and computation increase only when trends remain stable or change predictably. The talk uses experiments from Puranik’s work and frontier laboratories to examine data-mix questions that violate these assumptions, explain why small-scale results can mislead, and explore better methods for choosing training data. His research at Jane Street concerns scaling laws for sequence models in trading.

Ivo D. Dinov examines augmented intelligence as a partnership between people and computational systems. The talk traces the technical and philosophical development of neural networks, then introduces complex-time, or kime, representations and the inference methods they support. Applications include medical imaging, healthcare and economic forecasting. It also considers the academic mission, societal consequences of AI and tensions between rapid technical development and human values, framing mathematical AI methods as extensions of human reasoning. Free hybrid seminar hosted by Wayne State University's Institute for AI and Data Science, AIDaS: CAD Seminar Series. Wednesday 30 September 2026, 14:30–15:30 EDT (America/Detroit; UTC−4). Attend in Room 1146, Faculty/Administration Building, 656 W. Kirby, Detroit, MI 48202, USA, or use the Zoom link on the official event page. Researchers, practitioners, students and faculty are welcome; the organizer encourages in-person attendance.

Navid Azizan presents methods that make neural models obey physical, safety and operational constraints at deployment. Hard-constrained neural networks, or HardNets, enforce input-dependent constraints by construction while preserving universal approximation within the feasible function class. Applications include models of chaotic dynamics with bounded trajectories, energy-constrained operator learning, safe reinforcement learning and control with formal guarantees. The talk then turns to enforcing constraints during sampling from pretrained diffusion and flow-matching models. Formulating generation as trajectory optimization allows receding-horizon control to guide outputs toward feasibility without retraining or imposing excessive restrictions on the sampling process. Examples from fluid dynamics, robot planning and control, PDE control and language-guided image editing illustrate how constraints can define admissible behavior while retaining expressive learning and generation.

Lenka Zdeborová introduces the Rules-and-Facts model, in which some observations follow a shared rule while others are isolated exceptions requiring memorization. The framework studies when a learner can acquire the rule and retain those exceptions simultaneously. Its results emphasize how capacity is organized and deployed, rather than capacity alone. The talk examines how regularization and the geometry of kernels or learned feature maps can reserve resources for memorization without undermining rule learning. It connects the balance between abstraction and memory to architectural and algorithmic choices, providing a theoretical account of neural networks that both generalize and recall specific facts.

MIT Biology Colloquium Series talk by Peter Koo (Cold Spring Harbor Laboratory), hosted by Yunha Hwang, on using AI to learn the gene regulatory code.

Irene Rombel discusses how artificial intelligence and machine learning can address bottlenecks, quality issues, and production costs in cell, gene, and RNA therapy manufacturing. The seminar focuses on real-world process-development applications that improve reproducibility, speed, and cost effectiveness.

Florian Marquardt surveys approaches to reducing the energy and computational cost of neural networks by implementing their functions in physical devices. The central question is how experiments can efficiently determine updates to adjustable parameters, allowing the hardware itself to learn. Examples include learning through time reversal, obtaining nonlinear computation from linear optical scattering, and training analog quantum simulators as neuromorphic systems. The seminar is available in person and through the registration link on the organizer’s event page.

Qihao Ye examines practical, affordable ways for individual mathematicians to use AI in literature searches, proof development and verification. The talk identifies seven recurring failure modes and introduces the open-source QED multi-agent system designed to address them. A case study on constructing Carleman weights for wave equations on the half-line shows how automated search can be combined with a researcher’s mathematical insight and rigorous checking. The organizer links both the implementation and the accompanying research paper. Online via Rutgers DIMACS Zoom. Wednesday 23 September 2026, 10:00–11:00 EDT (America/New_York; UTC−4). Use the Zoom link and passcode on the official event page. Organized by the DIMACS AI and Mathematics Seminar. This listing advertises the verified remote attendance option.

ISTA Lecture by Klaus Robert Müller (TU Berlin & Korea University, Seoul) on how machine learning and AI enable scientific research, particularly in medicine and chemistry, and on explainability techniques for extracting understanding from machine learning models.

Omar M. Yaghi examines how reticular chemistry changes the scale and diversity of materials that chemists can design. By joining molecular building blocks into extended structures, researchers have developed large families of metal–organic frameworks and covalent organic frameworks with tunable internal spaces and functions. The lecture asks how artificial intelligence and robotics could accelerate the preparation and exploration of these materials, moving chemical discovery from scarcity toward abundance. Yaghi connects this approach to carbon capture and the collection of water from air, building on design principles for nanoporous frameworks and their applications in gas storage and separation. The discussion extends the metal–organic framework research recognised by the 2025 Nobel Prize in Chemistry into current methods for discovery and environmental applications.

Tal Linzen, an associate professor of linguistics and data science at New York University and a research scientist at Google, will speak in MIT's Siegel Family Quest for Intelligence seminar series. His work combines behavioral experiments and computational methods to study language learning and comprehension, alongside large-language-model post-training, evaluation, and interpretability.