Tommy Botch

Excited to be in Amsterdam for #CCN2025! Come learn about a project combining many of my favorite interests - naturalistic neuroimaging, development, & manifold learning (of course). Catch me at poster C181 on Fri - “Intrinsic dimensionality of brain activity manifolds across tasks and development”

Geometry of neural dynamics along the cortical attractor landscape reflects changes in attention The brain is a complex dynamical system whose activity reflects changes in internal states, such as attention. While prior work has shown that large-scale brain activity reflects attention, the mechan...

Preprint⭐
Our attention changes over time and differs across contexts—which is reflected in the brain🧠 Fitting a dynamical systems model to fMRI data, we find that the geometry of neural dynamics along the attractor landscape reflects such changes in attention!
www.biorxiv.org/content/10.1...

Computational Urban Ecology of New York City Rats Urban rats are highly adaptable, thriving in the dynamic and often inhospitable conditions of modern cities. Despite substantial mitigation efforts, they remain an enduring presence in urban environme...

New preprint!



tl;dr — We ran around late at night to record wild rats in NYC and figured out how to quantify their behavior and environment. 🧵

w/ Dima Batenkov, @zamakany.bsky.social, Emily Mackevicius

www.biorxiv.org/content/10.1...

What are the organizing dimensions of language processing?

We show that voxel responses during comprehension are organized along 2 main axes: processing difficulty & meaning abstractness—revealing an interpretable, topographic representational basis for language processing shared across individuals

There's a lot more in the paper (lookin' at you, prosody) -- we hope our work demonstrates the power of using LLMs as alternative models to understand language writ large! None of this would have been possible without the support of @esfinn.bsky.social and the FinnLab :)

How does sensory context facilitate language acquisition? We hypothesized that training LLMs with richer sensory context would partially ”rescue” performance relative to humans.

LLMs trained with richer sensory context performed better with much less data (~20%)!

However, LLMs are can still be useful models of behavior –LLM prediction distributions better reflected human predictions when reading written language.

Human predictions were more accurate than LLMs, even when just reading! This performance gap scaled with the amount of sensory context provided to humans.

We collected human predictions of natural language under varying sensory conditions (written, audio- only, or audiovisual language) and compared these with LLM predictions to the same stories.

Humans learn and use language informed by audiovisual information. In contrast, LLMs are often trained solely on written text – i.e., their view of language lacks the sensory richness of our human experience.

How essential are these sensory contexts for cultivating "human-like" language processing?

New preprint! Thrilled to share my latest work with @esfinn.bsky.social -- "Sensory context as a universal principle of language in humans and LLMs"

osf.io/preprints/ps...

Accelerated learning of a noninvasive human brain-computer interface via manifold geometry Brain-computer interfaces (BCIs) promise to restore and enhance a wide range of human capabilities. However, a barrier to the adoption of BCIs is how long it can take users to learn to control them. W...

New preprint! Excited to share our latest work “Accelerated learning of a noninvasive human brain-computer interface via manifold geometry” ft. outstanding former undergraduate Chandra Fincke, @glajoie.bsky.social, @krishnaswamylab.bsky.social, and @wutsaiyale.bsky.social's Nick Turk-Browne 1/8

Have you wondered where scene perception and memory intersect in the brain?

Do you want to localize these areas?

In a preprint, @carolinerobertson.bsky.social, Deepa Prasad, Brenda Garcia, and I detail this topography and release these parcels and our localizer.

www.biorxiv.org/content/10.1...

Thrilled to share new work w @taylorwwebb.bsky.social Shanka Mondal:
A Prefrontal Cortex-inspired Architecture for Planning in Large Language Models (arxiv.org)

LLMs struggle w multi-step planning.
We propose a solution inspired by brains: planning via recurrent interactions of PFC subregions.
🧵1/n

We review behavioral, computational, and neuroimaging evidence that converge to suggest that social interaction perception is visual and distinct from higher level social processes like theory of mind.

Functional networks in the infant brain during sleep and wake states Abstract. Functional brain networks are assessed differently earlier versus later in development: infants are almost universally scanned asleep, whereas adults

Resting state, or sleeping state? We know that wakefulness matters for adult functional brain network organization. In our new paper, we show that wakefulness can also be important in infant fMRI! 1/9

This project was a great way to start my PhD with @esfinn.bsky.social ! Thanks to everyone in the FINN Lab (and many others) for the help and support over the course of this work.

9/9

Why might this dissociation exist?

We suggest that concrete words share additional properties beyond purely linguistic information, such as imageability, that stems from integrating visual information into their neural representations.

8/9

While NLP models generally align with how humans represent language, we found that this is not always the case.

Neural representations of concrete concepts were more similar than those of abstract concepts, despite spanning larger distances in semantic space!

7/9

This result was surprising to us! Why do concrete concepts show more reliable representations than abstract concepts?

We used NLP tools to cluster concrete & abstract concepts in semantic space and compared neural representations of these “concept clusters” across stories.

6/9

Are concrete and abstract concepts contributing equally to this reliability?

By dichotomizing the continuous concrete-abstract axis, we show that concrete concepts exhibit more stable patterns of representation than abstract concepts.

5/9

These patterns of representations were not only reliable, but also unique to individual subjects.

Using a fingerprinting approach, we demonstrate the ability to identify subjects solely based on their representations of concrete and abstract concepts!

4/9

Across stories, we found that an axis of concrete and abstract concepts evoked highly reliable neural representations — more reliable than other semantic properties.

3/9

We compared the neural representations of semantic properties across diverse stories to identify properties of language that 1) varied across the population and 2) were reliable within individuals across experiences.

2/9

Neural representations of concrete concepts enable identification of individuals during naturalistic... bioRxiv - the preprint server for biology, operated by Cold Spring Harbor Laboratory, a research and educational institution

Why do people process language differently from each other? In a new paper, @esfinn.bsky.social and I show that the “concreteness” of word meanings plays a central role in shaping our unique experiences of natural language.

Preprint: www.biorxiv.org/content/10.1...

Thread below! 1/9