Search Results for author:
Found 20 papers, 6 papers with code
Generalization properties of contrastive world models
We systematically test the model under a number of different OOD generalization scenarios such as extrapolation to new object attributes, introducing new conjunctions or new attributes.
Inferring Inference
Given a time series of neural activity during a perceptual inference task, our framework finds (i) the neural representation of relevant latent variables, (ii) interactions between these variables that define the brain's internal model of the world, and (iii) message-functions specifying the inference algorithm.
Understanding robustness and generalization of artificial neural networks through Fourier masks
Our results also indicate that the essential frequencies in question are effectively the ones used to achieve generalization in the first place.
Learning Dynamics and Structure of Complex Systems Using Graph Neural Networks
In this work we trained graph neural networks to fit time series from an example nonlinear dynamical system, the belief propagation algorithm.
Interpolating between sampling and variational inference with infinite stochastic mixtures
This work takes a step towards a highly flexible yet simple family of inference methods that combines the complementary strengths of sampling and VI.
Phase transitions in when feedback is useful
We show that there is a non-monotonic dependence of optimal feedback gain as a function of both the computational parameters and the world dynamics, leading to phase transitions in whether feedback provides any utility in optimal inference under computational constraints.
Two-argument activation functions learn soft XOR operations like cortical neurons
Neurons in the brain are complex machines with distinct functional compartments that interact nonlinearly.
Generalization of graph network inferences in higher-order graphical models
A major challenge for these graphical models is that inferences such as marginalization are intractable for general graphs.
Exploring representation learning for flexible few-shot tasks
In this work, we consider a realistic setting where the relationship between examples can change from episode to episode depending on the task context, which is not given to the learner.
Neuromatch Academy: Teaching Computational Neuroscience with global accessibility
Neuromatch Academy designed and ran a fully online 3-week Computational Neuroscience summer school for 1757 students with 191 teaching assistants working in virtual inverted (or flipped) classrooms and on small group projects.
Probing Few-Shot Generalization with Attributes
Despite impressive progress in deep learning, generalizing far beyond the training distribution is an important open challenge.
Inverse Rational Control with Partially Observable Continuous Nonlinear Dynamics
This problem can be solved by control theory, which allows us to find the optimal actions for a given system dynamics and objective function.
Learning From Brains How to Regularize Machines
We propose to regularize CNNs using large-scale neuroscience data to learn more robust neural features in terms of representational similarity.
Inverse Rational Control with Partially Observable Continuous Nonlinear Dynamics
This approach provides a foundation for interpreting the behavioral and neural dynamics of highly adapted controllers in animal brains.
Improved memory in recurrent neural networks with sequential non-normal dynamics
In the presence of a non-linearity, orthogonal transformations no longer preserve norms, suggesting that alternative transformations might be better suited to non-linear networks.
Belief dynamics extraction
Estimating the structure of these internal states is crucial for understanding the neural basis of behavior.
Inverse Rational Control: Inferring What You Think from How You Forage
Complex behaviors are often driven by an internal model, which integrates sensory information over time and facilitates long-term planning.
Inference in Probabilistic Graphical Models by Graph Neural Networks
Message-passing algorithms, such as belief propagation, are a natural way to disseminate evidence amongst correlated variables while exploiting the graph structure, but these algorithms can struggle when the conditional dependency graphs contain loops.
Reviving and Improving Recurrent Back-Propagation
We examine all RBP variants along with BPTT and TBPTT in three different application domains: associative memory with continuous Hopfield networks, document classification in citation networks using graph neural networks and hyperparameter optimization for fully connected networks.
Skip Connections Eliminate Singularities
Here, we present a novel explanation for the benefits of skip connections in training very deep networks.
