Assistant Professor of Cognitive Science at Johns Hopkins. My lab studies human vision using cognitive neuroscience and machine learning.
Baltimore, MD
Joined January 2019
- Tweets276
- Following322
- Followers964
- Likes354
Pinned Tweet
Do neuroscientists need deep learning to explain brain representations? Our new paper shows that architectural manipulations alone—without pretraining—go surprisingly far in explaining visual cortex representations. Led by a fantastic former master’s student @AtlasKazemian
Super excited to share that my Master’s project, “Convolutional architectures are cortex-aligned de novo,” has been published in Nature Machine Intelligence! nature.com/articles/s42256-0… w/ @EricElmoznino @michaelfbonner
Mick Bonner retweeted
What is the nature of universal representations in AI models, and what determines whether they emerge?
Our paper accepted at #neurips2026 led by @florianmahner & @andropar addressed these questions comparing 162 vision models, with intriguing results.
arxiv.org/abs/2605.13675
👇
Mick Bonner retweeted
This paper is now published!
cell.com/iscience/fulltext/S…
And here is a news digest from Yahoo! Tech:
tech.yahoo.com/ai/articles/a…
New preprint from our lab!
Systematic image perturbations reveal persistent gaps between human and machine vision.
biorxiv.org/content/10.64898…
We compared humans and >200 DNNs in their visual cue usage during object recognition, and reveal systematic differences.
Mick Bonner retweeted
Excited to share our preprint! w/@martin_hebart
To understand how primates visually process objects in the world, we rely on both research in human and macaque IT. But what representations of object space are actually shared between them?
biorxiv.org/content/10.64898…
Quick thread🧵
Mick Bonner retweeted
Our review on superposition in AIs and brains is finally published in Nature Machine Intelligence 🙌
nature.com/articles/s42256-0…
Mick Bonner retweeted
1/ We’re so glad to share this new study 💫
Does the brain learn like a Deep Net? 🧠⚙️
- 📄Misalignment Between Backpropagation and the Hierarchy of Brain Responses to Images
- 🔗arxiv.org/abs/2605.28693
Thread below 🧵
Mick Bonner retweeted
We've updated the preprint of our Naturalistic Computational Cognitive Science paper — we've clarified and streamlined the arguments, and expanded examples where we see increasing naturalism already yielding new theoretical insights, from RL to perceptual neuroscience. 1/4
ALT Two examples of naturalistic computational cognitive science. Given a computational theory, researchers generate initial predictions to test their theory in a simplified setting. Theories are validated by their ability to predict human behavior in increasingly naturalistic conditions, and refined where they fail to predict correctly. (Top) Progression of theories of sequential task generalization. (Bottom) Examples of progression in understanding PRC involvement in difficult visual judgments..
Mick Bonner retweeted
Our NeuroAI study made it onto the cover of Nature Machine Intelligence (@NatMachIntell) ❤️.
In it, we demonstrate that a developmentally-inspired visual diet can drastically improve the robustness of ANN-based vision systems.
open access, open code, open weights, open science
Mick Bonner retweeted
We are excited to open the re:vision initiative, a community-driven initiative for replicating and generalizing findings in visual neuroscience, based on the LAION-fMRI dataset.
re-vision-initiative.org/
Why this initiative and why would you want to participate? 🧵
Mick Bonner retweeted
Are you attending #VSS2026?
Come check out my talk on cross-species alignment for finding shared and distinct representational geometries in primate IT.
Saturday, May 16, 2026, 3.30pm, Talk Room 1
Mick Bonner retweeted
Very excited about #VSS2026 starting today! We hope you can make it to our lab’s presentations! Come and visit us at the talks & posters, and make sure you come to our workshop about the community replication initiative re:vision. Low bar for entry! Details will follow!
Excited for this! I’ll be there, and will be giving a talk.
Happy to announce the third iteration of NEAT (Neuro-AI-Talks), which will take place in Osnabrück September 14th-15th 2026.
NEAT is a (deliberately small scale) NeuroAI workshop that brings together researchers from neuroscience and AI.
kietzmannlab.org/neat2026/
More info👇
Mick Bonner retweeted
Congratulations Cliona, and the rest of the FOUNDCOG team, I am so excited to see these results published!
Mick Bonner retweeted
1/7 Does the infant brain have representational structure? 👶🧠In the FOUNDCOG project, we scanned 134 awake infants using fMRI. Published in Nature Neuroscience, our research reveals 2-month-old infants already possess complex visual representations in VVC that align with DNNs.
New paper from our lab on the behavioral significance of high-dimensional neural representations!
Human visual cortex representations may be much higher-dimensional than earlier work suggested, but are these higher dimensions of cortical activity actually relevant to behavior? Our new paper tackles this. 🧵(1/n) cell.com/current-biology/ful…
Mick Bonner retweeted
Great profile on @minyoung_huh, @thisismyhat, @ssnl_tz & @phillip_isola. Pairs well our work with @ev_fedorenko on Brain/ANN convergence in language & vision (tinyurl.com/4d3p53hb) & @ziruichen44 & @michaelfbonner’s work on visual reps (tinyurl.com/4d3p53hb)
Recently, a team of AI researchers found that different models can develop similar internal representations, even if they’re trained on entirely different data types.
quantamagazine.org/distinct-…
Dimensionality reduction may be the wrong approach to understanding neural representations. Our new paper shows that across human visual cortex, dimensionality is unbounded and scales with dataset size—we show this across nearly four orders of magnitude. journals.plos.org/ploscompbi…
Why did so many previous studies report low dimensionality? 1. High-quality neural datasets are finally large enough to probe representations beyond just tens of dimensions! 2. Standard methods in cognitive neuroscience are insensitive to low-variance—but meaningful—dimensions.