N. Apurva Ratan Murty

Braintree: An agentic framework for in-silico vision neuroscience

Braintree autonomously runs in-silico vision neuroscience experiments to test hypotheses, generate stimuli, and analyze neural responses using brain models.

Standardized human tests show that most vision models are face-blind

The paper tests AI models on standardized face recognition tasks and finds most perform worse than humans, with some showing face-blindness similar to human prosopagnosia.

Zero-shot evaluations expose structured generalization limits in predictive models of the brain

The paper introduces a zero-shot evaluation framework to assess predictive models of the brain, revealing structured generalization limits and linking prediction performance to cognitive neuroscience findings.

Category selectivity observed in the human brain is distinct from category selectivity observed in artificial neural networks

The paper compares category selectivity in human brain regions with that in artificial neural networks, finding significant differences in response tuning and stability.

Individual differences in artificial neural networks capture individual differences in human behavior

This study shows that individual differences in artificial neural networks mirror human behavioral differences in accuracy, confidence, and response time across tasks.

Can category-selective cortex predict categorisation behaviour?

The paper shows that activation in category-selective regions of the occipito-temporal cortex predicts categorisation behaviour in face, body, and scene classification tasks.

MOSAIC: A scalable framework for fMRI dataset aggregation and modeling of human vision

MOSAIC aggregates eight fMRI datasets to enable large-scale modeling of human vision, showing improved decoding accuracy with diverse stimuli and joint brain-optimized encoding models.

Human-like individual differences emerge from random weight initializations in neural networks

This study shows that different random initializations of neural networks produce individual differences similar to humans, and using these differences improves predictions of human responses.

Visual experience contributes to separation of face and language responses in the ventral stream

The paper shows that visual experience helps separate face and language responses in the ventral stream, while blind individuals show overlapping responses.

BOLD Moments: modeling short visual events through a video fMRI dataset and metadata

The paper introduces BOLD Moments Dataset (BMD), a video fMRI dataset capturing whole-brain responses to naturalistic videos, and shows its ability to capture temporal dynamics and correlate with video memorability.

Fine-grained neural coding of bodies and body parts in human visual cortex

This study records multi-unit spiking activity in human visual cortex to investigate body and body part representation, finding strong selectivity for bodies and generalization to non-photographic depictions.

A Highly Selective Response to Food in Human Visual Cortex Revealed by Hypothesis-Free Voxel Decomposition

This study uses data-driven voxel decomposition to identify neural selectivities in human visual cortex, revealing a strong food-selective component not previously identified.

Using child-friendly movie stimuli to study the development of face, place, and object regions from age 3 to 12 years

This study uses child-friendly movies to identify age-related changes in brain regions selective for faces, scenes, and objects from age 3 to 12 years.

Visual Experience is not Necessary for the Development of Face Selectivity in the Lateral Fusiform Gyrus

The paper shows face-selectivity in the lateral fusiform gyrus of congenitally blind individuals through haptic exploration, indicating visual experience is not required for this selectivity.