Interpreting neural computations by examining intrinsic and embedding dimensionality of neural activity

2021/09/17 by Mehrdad Jazayeri, Srdjan Ostojic · 1 voice · 38 citations
Neuroscience · #Neural dynamics and brain function #Functional Brain Connectivity Studies #Visual perception and processing mechanisms

paper · pdf · doi:10.1016/j.conb.2021.08.002

Cited by

• From sensory to perceptual manifolds: The twist of neural geometry
• The limits of falsifiability: Dimensionality, measurement thresholds, and the sub-Landauer domain in biological systems
• Discovering cognitive strategies with tiny recurrent neural networks
• Characterizing Neural Manifolds' Properties and Curvatures using Normalizing Flows
• Characterizing higher-order representations through generative diffusion models explains human decoded neurofeedback performance
• Voltage imaging as a window into neural computation
• Shaping manifolds in equivariant recurrent neural networks
• Rarely categorical, highly separable representations along the cortical hierarchy
• Quantifying the compressibility of the human brain
• Universal scale-free representations in human visual cortex
• Category learning in deep neural networks: Information content and geometry of internal representations
• Statistics of correlations in nonlinear recurrent neural networks
• A complex network perspective on brain disease
• Fluorescence imaging of large-scale neural ensemble dynamics. [europepmc]
• Orthogonal representations for robust context-dependent task performance in brains and neural networks. [europepmc]
• Theory of the Multiregional Neocortex: Large-Scale Neural Dynamics and Distributed Cognition. [europepmc]
• Cognition and the single neuron: How cell types construct the dynamic computations of frontal cortex. [europepmc]
• The role of conjunctive representations in prioritizing and selecting planned actions. [europepmc]
• Recent Advances at the Interface of Neuroscience and Artificial Neural Networks. [europepmc]
• Neural manifold analysis of brain circuit dynamics in health and disease. [europepmc]
• Parametric control of flexible timing through low-dimensional neural manifolds. [europepmc]
• Geometric transformation of cognitive maps for generalization across hippocampal-prefrontal circuits. [europepmc]
• A unifying perspective on neural manifolds and circuits for cognition. [europepmc]
• Learnable latent embeddings for joint behavioural and neural analysis. [europepmc]
• Multi-view manifold learning of human brain-state trajectories. [europepmc]
• Dynamical flexible inference of nonlinear latent factors and structures in neural population activity. [europepmc]
• High-performing neural network models of visual cortex benefit from high latent dimensionality. [europepmc]
• Simultaneous, cortex-wide dynamics of up to 1 million neurons reveal unbounded scaling of dimensionality with neuron number. [europepmc]
• Dimensionality reduction beyond neural subspaces with slice tensor component analysis. [europepmc]
• Thalamocortical architectures for flexible cognition and efficient learning. [europepmc]
• Neural representational geometries reflect behavioral differences in monkeys and recurrent neural networks. [europepmc]
• Semi-orthogonal subspaces for value mediate a binding and generalization trade-off. [europepmc]
• A transient high-dimensional geometry affords stable conjunctive subspaces for efficient action selection. [europepmc]
• Decoding the brain: From neural representations to mechanistic models. [europepmc]
• The unbearable slowness of being: Why do we live at 10 bits/s? [europepmc]
• The dynamics and geometry of choice in the premotor cortex. [europepmc]
• Discovering cognitive strategies with tiny recurrent neural networks. [europepmc]
• Neural population geometry: An approach for understanding biological and artificial neural networks. [europepmc]

Discussions

• It feels like the converastion about manifold dimensionality is back, so I thought I'd share a paper that explains nicely why measuring the "embedding dimensionality" of a manifold (e.g., counting PCs [bsky, 39 points, 1 comments]