Bayesian inference

…Participants implement likelihood and Bayesian inference, including Markov chain Monte Carlo, and study Brownian motion, continuous-time Markov chains and birth–death processes. These tools are applied to trait evolution, biogeography and diversification, including the effects of environmental and biotic interactions. The practical course is intended for researchers familiar with programming, probability and phylogenetics; prior Julia experience is not required. Sessions on 1, 3, 5, 8, 10 and 12 March 2

…uncertainty quantification, Bayesian inference and physics-informed neural networks. Build scalable simulation software and apply it with experimental and clinical collaborators to development and disease models.

…regulation using modern deep learning, Bayesian inference and causal modelling. Work at the interface of computational research and large-scale biological data.

…complex ways for modeling and inference. Meanwhile, capturing the dependency on experimental conditions surrounding the neural recordings remains understudied, and few methods leverage the correlation structure across experimental conditions. This poses difficulties for model estimation as well as any analyses comparing and interpolating between conditions. In this work, we develop the Gaussian Process Linear Dy

Neural circuits display both input-driven activity that is necessary for the real-time control of behavior and internally generated activity that is necessary for memory, planning, and other cognitive processes. A key mediator between these intrinsic and evoked dynamics is arousal, an internal state variable that determines an animal's level of engagement with its environment. In this work, we show that a continuous Hopfield network embellished with a single additional parameter---recurrent gain---captures some essential effects of arousal state at neural and cognitive levels. Using the model

Recent evidence suggests the brain performs sampling-based Bayesian inference to solve various cognitive tasks [1–6]. Neural sampling requires neural circuits to exhibit statistically stable variability to draw posterior samples. While ad-hoc frameworks simply inject randomness into spike emission, it remains unknown how such variability for sampling is generated in biological neural circuits. Here we propose that recurrent spiking circuits can internally generate statistically stable variability for posterior sampling, eliminating the need for an external random source. Our proposed circuit m

To survive, organisms must make rapid inferences about their surroundings. For example, to successfully escape approaching predators, prey must estimate the direction of approach from noisy stimuli [1]. Such rapid inferences are particularly challenging because there is little time to gather sensory information, yet the precision of inference is critical for survival. Due to evolutionary pressures, nervous systems likely evolved effective computational strategies that enable accurate inferences under strong time limitations. Here, we aim to understand the principles underlying such rapid infe

Hippocampal remapping describes the phenomenon of substantial reshuffling of the spatial preferences of hippocampal place cells in response to environmental changes. Disparate neural representations in different environments could reduce interference between competing memories, hence functionally supporting robust context-dependent spatial cognition. Experimental data on existing remapping studies exhibit a diverse range of physiological properties, sometimes with seemingly conflicting results. In particular, subtle changes in experimental conditions lead to qualitatively different results. De

This webinar addressed computational perspectives on how animals and humans make decisions, spanning normative, descriptive, and mechanistic models. Sam Gershman (Harvard) presented a capacity-limited reinforcement learning framework in which policies are compressed under an information bottleneck constraint. This approach predicts pervasive perseveration, stimulus‐independent “default” actions, and trade-offs between complexity and reward. Such policy compression reconciles observed action stochasticity and response time patterns with an optimal balance between learning capacity and performan

…be understood through the framework of Bayesian inference. We find that separability only depends on three geometric properties of the neuronal population vector, described by cones within th

Sensory anomaly in autism is thought to arise from atypical Bayesian inference processes in the brain. It remains elusive, though, which neuronal circuits are involved. Here we focused on the superior colliculus and performed in vivo two-photon calcium imaging while the mice were engaged in an implicit visual learning task. Following the Bayesian learning principles, wild-type mice showed contextual modulation of the visual responses in the superior colliculus and pupil dynamics, but not in the retinal signals. In contrast, Scn2a-haploinsufficient autism model mice exhibited anomalous modulati

…We compare several modern Bayesian inference schemes for model inversion on simulated data . We benchmark the identifiability of local parameters governing the interplay between several neuronal populations in each region, as well as global parameters t

…In this study our aim is to conduct inference and prediction on the local and global dynamics using virtual brain models, affected by disorders. Inference algorithms are necessary to efficiently estimate the unknown parameters (such as regional excitability parameters and global scaling factor on connectome), ideally incorporating measures of uncertainty. In this work, we provide f

…Bayesian inference postulates the use of perceptual priors to increase perceptual acuity in the face of uncertainty. Importantly, dynamic environments require a flexible use of priors to enable quick adaptation to environmental changes.The present study investigates Bayesian perceptual adaptation in dynamic environments through EEG and pupillometry. We had 26 young adults perform an auditory motion discrimination task, indicating the final directions of auditory motion sequences with ran

Predicting the location of expected upcoming stimuli is a key function in sensory processing. Noise in the environment reduces the reliability of sensory input and increases the importance of prior information. Yet, sudden changes in the environment can render this prior information irrelevant. In such cases, observers should reset their beliefs and establish new priors using current sensory evidence. To study these dynamics of perceptual belief updating, we conducted a human EEG experiment (N = 29) investigating auditory spatial localization in noisy volatile environments and analyzed the ass

…Bayesian inference prescribes a statistically optimal solution to this problem, but it is memory intensive and computationally complex, thus unlikely to be used continuously by humans.Here, we systematically investigated which hallmarks of Bayesian inference are present in previously published human response

Predicting the timing of external stimuli is a key function in sensory processing. In noisy environments, the importance of prior information increases due to reduced reliability of sensory information. At the same time, sudden changes in the environment often render prior information invalid. In such cases, observers should update beliefs and establish new priors using current sensory evidence. To study these dynamic processes, our current EEG experiment investigates the behavioral and neural patterns associated with auditory temporal discrimination in a noisy and dynamically changing environ

The large diversity of behaviors even among closely related species indicates the evolvability of the underlying neural circuit. At the same time, the behaviors must be functionally robust, but how systems can be both robust and evolvable is still an open question in neuroscience and evolution. Studying robustness and evolvability requires the mapping between genotypes and phenotypes, which is challenging to obtain experimentally. However, models of neural circuits that generate behavior can be used as a proxy of the biological system, and the mapping between model parameters and model output

Over the past decade we have demonstrated that the fusion of subject-specific structural information of the human brain with mathematical dynamic models allows building biologically realistic brain network models, which have a predictive value, beyond the explanatory power of each approach independently. The network nodes hold neural population models, which are derived using mean field techniques from statistical physics expressing ensemble activity via collective variables. Our hybrid approach fuses data-driven with forward-modeling-based techniques and has been successfully applied to expla

Over the past decade we have demonstrated that the fusion of subject-specific structural information of the human brain with mathematical dynamic models allows building biologically realistic brain network models, which have a predictive value, beyond the explanatory power of each approach independently. The network nodes hold neural population models, which are derived using mean field techniques from statistical physics expressing ensemble activity via collective variables. Our hybrid approach fuses data-driven with forward-modeling-based techniques and has been successfully applied to expla

…The framework of Bayesian inference successfully formalizes this knowledge in terms of probabilistic priors, and perception/action as probabilistic inference. However, our understanding of how such priors are obtained, represented and used by the brain remains limited. Here we build upon the recent success in machine learning of “diffusion models” as a means of learning and usin

…of perception (efficient coding, Bayesian inference) that are obscured when studied in

Aggression, leading to the formation of social hierarchy, has an evolutionary role in reducing conflict and facilitating the allocation of limited resources. Previous studies in mice have identified the regions involved in the neural circuitry of aggression, yet the mechanisms leading to the emergence of aggressive behavior remain to be poorly understood. To analyze such mechanisms, we used a mouse chronic social conflict paradigm where multiple aggressive phenotypes were formed, for which we obtained behavioral readouts (134 mice) and whole-brain snapshots of neuronal activity (c-Fos; 54 mice

…from basing their decisions on smart inferences about hidden properties of the world. Typical theoretical approaches for understanding the strategies that animals use in such settings combine Bayesian inference and value iteration to derive optimal behavioral policies that maximize total reward given changing beliefs about the environment. However, specifying these beliefs requires infinite numerical precision; with limited resources, this problem can no longer be decomposed into the separate steps of optimizing inference