Talia Konkle

The paper investigates coarse-scale functional organization in occipitotemporal cortex using fMRI data, revealing complexities in animacy-size organization and alternative dimensions.

The paper shows that leading neural network models of primate visual cortex differ in their ability to control neural firing using parametric stimulus perturbations, revealing discrepancies in how models and neurons parameterize visual space.

This paper uses deep learning to show that lightness illusions can emerge from edge-based image reconstruction, suggesting a simpler computational mechanism than previously thought.

This paper shows that feedforward CNNs can mimic human contour integration without lateral connections, recurrence, or feedback by leveraging receptive field progression and contour bias.

This paper shows that large language models can predict visual brain responses in humans and non-human primates, suggesting that language-aligned models capture general visual patterns rather than language-specific features.

This paper introduces contrastive coding as a framework explaining category-selective regions in the human brain using neural network analyses.

This study develops an ultra-wide angle neuroimaging method to investigate how immersive scenes are represented in the brain, finding that scene regions show minimal modulation to visual size changes and maintain content preferences even with central visual field loss.

This paper demonstrates that a simple synaptic growth rule generates modular, hierarchical, and topographically organized visual pathways in a developmental model.

The paper shows that face cells in macaques respond to non-faces in a way that predicts face selectivity, challenging the idea that face cells are exclusively face-specific.

The paper shows that early visual cortex encodes signals for high-level object categorization, challenging the idea that only higher visual areas are involved.

The paper investigates how the brain encodes spatial scale in images, finding that while individual voxels show ramp-shaped tuning, multi-voxel patterns in the ventral occipito-temporal cortex represent the object-to-scene continuum smoothly and linearly.

The paper investigates whether letter perception relies on specialized letter features or general object-based features, finding that general object-based features better account for perceptual similarity.

The paper uses self-organizing principles to show that a unified representational space in a deep neural network can produce brain-like topographic motifs similar to the object-responsive cortex.

Across 224 vision models scored against fMRI responses, architecture and training objective barely mattered; the images a model was trained on mattered most, and linear re-weighting blurs differences.

This paper presents a self-supervised model that learns visual representations of individual images, showing category information emerges implicitly and matches category-supervised models in capturing brain response structures.

The paper uses EEG to show that mid-level features in images rapidly activate neural responses for animacy and object size, even when images are unrecognizable.

The paper investigates how intuitive action understanding is based on inferred goals and movements, finding that goals predict similarity judgments best, with visual appearance playing a minor role.

This paper introduces a self-supervised model that learns instance-level representations of images, achieving performance comparable to category-supervised models in predicting human ventral visual stream responses.

The paper examines how neural responses to inanimate objects are organized by size and motor-relevance, finding that manipulability and stability explain responses in different brain regions.

This paper presents a model showing multiple mirrored visual areas emerge from spatial mapping without requiring hierarchical processing.

The paper identifies a distinct neural representation for reachable-scale environments in the brain, separate from object and scene views.

The paper investigates how the visual cortex organizes representations of human actions by identifying five networks related to social aspects and interaction envelope scale.

This paper introduces a voxel selection method for fMRI data that prioritizes reliability over activity, resulting in a smaller, more reliable set of voxels compared to existing methods.

The paper investigates how mid-level features explain large-scale organization in the ventral stream using novel stimuli called texforms.