Deep Learning

This Keystone Symposium brings structural and cellular biologists together to integrate cryo-electron microscopy, super-resolution and multiscale imaging, data science, computational modeling and deep learning. The programme connects molecular structure to tissue-level function and includes in-person, livestream and on-demand participation.

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.

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.

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.

Useful internal representations should explain the patterns of regularities and dependencies among observations. Probabilistic graphical models promise a principled way to uncover latent factors as such, but they are hard to scale to handle high-dimensional sensory observations and complicated dependencies structures. Neural-networks, on the other hand, excel at approximating complicated high-dimensional functions, but their internal representations do not easily lend themselves to a probabilistic interpretation. Despite some successes, a general unified approach is still missing for integrating the two approaches. I will describe a novel approach towards merging adaptive neural-network components into a probabilistic framework, based on three core ideas. The first is to train a set of networks to collectively perform inference, leveraging the ability of pattern-recognition methods to amortise complicated transformations. The second is to constrain the way in which the outputs of these networks are interpreted, transformed, and combined together. These constraints, together with the learning objective itself, are derived directly from probabilistic considerations encoded in a graphical model. Finally, the third core idea is that of recognition-parametrisation, allowing the inference ("recognition") procedure to directly define the model itself, without requiring an explicit "generative" decoder. Presented in the van Vreeswijk Theoretical Neuroscience Seminar series (formerly WWTNS) on 2026-03-04. Recording duration: 00:48:26.

1. Can we reconstruct images that a person saw, directly from their fMRI brain recordings? 2. Can we reconstruct the training data that a deep-network trained on, directly from the parameters of the network? The answer to both of these intriguing questions is “Yes!” In this talk I will present some of our work in both domains. I will then show how combining the power of Brains and Machines can lead to significant breakthroughs in both areas, and potentially bridge the gap between Minds and Machines. Finally, I will show how combining the power of Multiple Brains (with NO shared data) may lead to new breakthrough discoveries in Brain-Science, and allow mapping of information between different brains. Presented in the van Vreeswijk Theoretical Neuroscience Seminar series (formerly WWTNS) on 2026-02-18. Recording duration: 00:50:24.

We introduce a statistical physics framework for learning in neural architectures composed of single or interconnected asymmetric attractor networks. These systems can exhibit a manifold of global fixed points capable of implementing sophisticated input-output mappings, which we characterize analytically. Learning from extensive datasets is achieved through the stabilization of fixed points via a fully distributed and local learning process, implemented at the single-neuron level. This simple mechanism yields performance comparable to that of conventional feedforward deep neural networks trained using gradient-based methods. The effectiveness of the model stems from the dense and accessible manifolds of stable fixed points, which encode the internal representations of data. Unlike other approaches to deep learning without backpropagation, our method does not attempt to estimate gradients. Presented in the van Vreeswijk Theoretical Neuroscience Seminar series (formerly WWTNS) on 2025-06-18. Recording duration: 00:45:35.

Over the past decade, neuroscience, cognitive science and computer science (“AI”) converged to create specific, image-computable, deep neural network models intended to appropriately abstract, emulate and explain the mechanisms of primate ventral visual processing, up to its deepest neural level, the inferior temporal cortex (IT). Because these leading neuroscientific emulation models — aka “digital twins” — are fully observable and machine-executable, they offer predictive and potential application power that our field’s prior conceptual models did not. Our team’s ongoing work is aimed at asking if current digital twin models might support non-invasive, beneficial brain modulation. In this talk, I will describe a key result: we demonstrate that we can use a digital twin to design spatial patterns of light energy that, when “added” to the organism’s retinal input in the context of ongoing natural visual processing, results in precise modulation (i.e. rate bias) of the pattern of a population of IT neurons (where any intended modulation pattern is chosen ahead of time by the scientist). Because the IT visual neural populations are known to directly connect to and modulate downstream neural circuits (e.g. amygdala) that may underlie psychological affective states (e.g. mood and anxiety), this novel basic science may unlock a new, non-invasive application avenue of potential future human clinical benefit. This progress and new impact possibilities resulted from convergent brain science and AI engineering efforts in the domain of visual object intelligence. I will motivate this as just one example of what I believe will unlock in other domains of human intelligence as brain scientists and AI engineers collaborate to develop machine-executable models of the underlying mechanisms of those still-mysterious domains. CARL VAN VREESWIJK MEMORIAL LECTURE 2025. Presented in the van Vreeswijk Theoretical Neuroscience Seminar series (formerly WWTNS) on 2025-04-09. Recording duration: 00:55:49.

At Recod.ai, Unicamp, develop algorithms for continuous health monitoring with wearable devices and coordinate research with academic and industry collaborators. Apply through the linked online form with a cover letter and CV by 5 October 2026. The host specifies at least 40 hours per week.

Two-year, full-time postdoctoral appointment investigating commonsense reasoning in multimodal world models. Research combines neural and symbolic methods to represent abstractions, causality, space and time, with the aim of improving human-centred AI. The project involves collaboration across AI, computer vision and reasoning, alongside limited teaching and student supervision. The preferred start is in the first quarter of 2027.

Develop ARCA, a foundation model that combines crop microbiome, genome and environmental data to predict microbial community behaviour and support crop resilience. The NOAH consortium connects model development with greenhouse and field experiments. This 36–40-hour postdoctoral appointment starts in January 2027, initially for one year with a three-year extension after positive evaluation. Apply with a CV, motivation letter, two referee contacts and evidence of PhD completion or a planned defence.

Fully funded doctoral studentship in the Ali Lab at the Cancer Research UK Cambridge Institute, starting in October 2027. The project combines spatial multiomics with deep learning to identify treatment-response biomarkers in breast cancer. It will analyse large multimodal datasets from observational cohorts and clinical trials, using representation and cross-modal learning to study target expression, tissue architecture and tumour heterogeneity. Training spans quantitative pathology, multiplexed imaging and predictive modelling.

Fully funded doctoral studentship in the Ali Lab at the Cancer Research UK Cambridge Institute, starting in October 2027. The project combines spatial multiomics with deep learning to identify treatment-response biomarkers in breast cancer. It will analyse large multimodal datasets from observational cohorts and clinical trials, using representation and cross-modal learning to study target expression, tissue architecture and tumour heterogeneity. Training spans quantitative pathology, multiplexed imaging and predictive modelling.

Two-year, full-time postdoctoral appointment investigating commonsense reasoning in multimodal world models. Research combines neural and symbolic methods to represent abstractions, causality, space and time, with the aim of improving human-centred AI. The project involves collaboration across AI, computer vision and reasoning, alongside limited teaching and student supervision. The preferred start is in the first quarter of 2027.

Four-month, full-time fellowship for empirical AI safety and security research, with researcher mentorship, a weekly stipend and research support. Workstreams include model interpretability, scalable oversight, adversarial robustness and evaluation of advanced AI systems. Applications for the January cohort are open and reviewed on a rolling basis. Shared workspaces are available in Berkeley and London, with remote participation possible in the United States, United Kingdom or Canada. A subsequent permanent role is not guaranteed.

Four-month, full-time research fellowship developing machine-learning infrastructure and reinforcement-learning methods. Projects can address training performance, distributed systems, synthetic-data and environment pipelines, or model generalization. Fellows receive researcher mentorship, a weekly stipend and benefits. Applications are reviewed on a rolling basis for the next cohort, expected in January 2027. San Francisco workspace access and remote participation from the United States, United Kingdom or Canada are available. A subsequent permanent role is not guaranteed.