Generative models

…Topics include foundation models, cross-modality learning, causal explanation, mechanistic interpretability, generative models and AI agents. Onsite workshops and tutorials take place on the first day; the following three days offer hybrid talks and posters. Registration is not yet open; the official page accepts expressions of interest.

Biwei Dai explores generative models built around physical symmetries and multiscale structure. For cosmology, these models enable field-level likelihoods that extract more information from weak lensing than two-point statistics and expose model misspecification through generated samples. Related methods address lattice sampling and collider anomalies. Physically informed generative priors also support inverse problems, including intensity-interferometry reconstruction of black-hole accretion disks and analysis of pulsar-timing gravitational-wave backgrounds.

…talk, we discuss the development of generative models of human gaze in response to visual input. We discuss how such models can be learned, both using supervised learning and using implicit feedback as an agent interacts with the environment, the latter being more plausible in biological agents. We also discuss two ways such models can be used. Firs

Generative models provide a powerful interpretive framework in neural reinforcement learning. However, given the complexity of behavior, models are often proposed in the absence of structurally different alternatives, poten-tially weakening their explanatory force. We employ a data-driven methodology (choice-wide behavioral associa-tion study; CBAS; Kastner et al. BioRxiv 2024), to examine such limitations. We use a dataset of human subjects performing a two-step task that was designed to test the interplay between model-based and model free learning (Gillan et al. eLife 2016). The subjects al

…Hence, their performance as generative models tends to be subpar. To construct a generative model that entirely learns the statistics of neural activity with no a

…a research framework that uses generative models to learn the complex nonlinear mapping between visual attributes and cognitive properties (Goetschalckx et al., 2019). This approach identifies transformations in the latent space of a generative model, which, when applied, generate images with tailored visual attributes that can enhance or diminish targeted cognitive

…arise from a trio of domain-specific generative models (Fig 1a) within prefrontal (PFC) nodes of the Default-Mode Network (DMN), and integrate within its parietal nodes, forming the basis for our dynamic consciousness.Smoothed binary ratings of prediction updates for State changes, Agent beliefs, and Action sequences from an online sample were used as parametric regressors for the GLMs. These ratings indicating experience shifts

…of my talk, I will focus on diffusion models as generative mo

…of my talk, I will focus on diffusion models as generative mo

The structure and function of biological neural networks are the result of biological “construction rules”, physical limits, stochasticity, developmental noise, and learning. The nature of these building rules and the resulting architectures are fundamental to our understanding of neuro-developmental plans and the structure and function of neural circuits. The detailed reconstruction of connectomes makes it possible to study their design; Recent reconstruction of connectomes during different developmental stages in C. elegans allows for characterization of their development. We present a class

Elucidating the neural basis of cognition requires incorporating cognitive models into the modelling of neural data. A prevalent strategy in neuroimaging is to fit computational models of cognition to concurrent behavior, and then regress the latent variables of those models onto fMRI data. Though widely used, this approach is restricted to mapping computational variables that are in some way expressed in behavior. Here we introduce a computational parametric mapping (CPM), a framework which builds on and generalizes the Bayesian population receptive field framework. CPM offers three main adva

3D Morphable Models are my favorite class of generative models and are commonly used to model faces. They are typically applied to ill-posed problems such as 3D reconstruction from 2D data. I'll start my presentation with an introduction into 3D Morphable Models and show what they are capable of doing. I'll then focus on our recent finding, the Identity-Expression Ambiguity: We demonstrate that non-orthogonality of the variation in identity and expression can cause identity-expression ambiguity in 3D Morphable Models, and that in practice expression and identity are far from orthogonal and can

This talk will focus on the generative modelling of resting state time series or endogenous neuronal activity. I will survey developments in modelling distributed neuronal fluctuations – spectral dynamic causal modelling (DCM) for functional MRI – and how this modelling rests upon functional connectivity. The dynamics of brain connectivity has recently attracted a lot of attention among brain mappers. I will also show a novel method to identify dynamic effective connectivity using spectral DCM. Further, I will summarise the development of the next generation of DCMs towards large-scale, whole-

…of the brain has emerged have used generative mathematical models, which grow synthetic networks according to specific wiring rules. Evidence suggests that a wiring rule which emulates a trade-off between connection costs and functional benefits can produce networks that capture essential topological properties of brain networks. In this webinar, Professor