Language processing in Brains vs Machines PhD student @georgiatech
Atlanta
Joined October 2014
- Tweets656
- Following865
- Followers497
- Likes3K
Taha Binhuraib 🦉 retweeted
Thanks @GeorgiaTech for featuring @EyasAyesh and our ongoing work on detecting AI misalignment.
Preprint soon! (we decided to wait for reviewer comments first 😊)
news.gatech.edu/news/2026/10…
Come say hi! Lots of exciting work from our lab this year.
#CCN2026
The language, intelligence & thought (LIT) lab is presenting at #CCN2026 on Tuesday!
1/
ALT LIT lab @ CCN 2026 Kamya Hari - Independent-component-based encoding models of brain activity during story comprehension Poster B41 Taha Binhuraib - Divergent scaling patterns in the auditory cortex vs. other brain regions reveal distinct sources of LLM-brain alignment Poster B78 Anya Ivanova - Is NeuroAI adopting the right methods and theoretical frameworks? GAC
Taha Binhuraib 🦉 retweeted
Machines see the world fundamentally differently from humans, and it's costing us a lot of tokens.
Where machines see uniform rectangles, our brain sees a variable resolution retinal array reformatted into a magnified manifold in primary visual cortex, enabling high-resolution and wide field-of-view at minimal computational cost.
In our new ICML paper, we enable the same mechanism for deep vision models.
🧵 ↓
Join our lab!
Language, Intelligence & Thought lab is looking for a lab manager! This is a 2-year postbac position that will allow you to gain experience in human neuroscience, cognitive science, and AI research prior to applying to PhD programs.
Express interest here: forms.gle/289sLgZdJ2bQr1Y48
Taha Binhuraib 🦉 retweeted
This demonstrates that authors are not reading their own papers before submission. After all, reference errors are easy to catch. Thus it suggests tons of errors also in the math, methods, results, ...
Sad to see human slop being exacerbated by AI slop.
Okay so, we just found that over 50 papers published at @Neurips 2025 have AI hallucinations
I don't think people realize how bad the slop is right now
It's not just that researchers from @GoogleDeepMind, @Meta, @MIT, @Cambridge_Uni are using AI - they allowed LLMs to generate hallucinations in their papers and didn't notice at all.
It's insane that these made it through peer review👇
Watching @rogerfederer feels like hearing a language we no longer speak. Tennis once had rhythm, hesitation, and character. Now it's optimized, mechanical, and stripped of surprise. Precision replaced soul.
Taha Binhuraib 🦉 retweeted
What if all AI models share a hidden, low-dimensional "brain"?
Johns Hopkins University reveals that neural networks, regardless of task or domain, converge to remarkably similar internal structures.
Their analysis of 1,100+ models (Mistral, ViT, LLaMA) shows they all use a few key "spectral directions" to store information.
This universal structure outperforms assumptions of randomness, offering a blueprint for more efficient multi-task learning, model merging, and drastically cutting AI's computational and environmental costs.
The Universal Weight Subspace Hypothesis
Paper: arxiv.org/pdf/2512.05117
Page: toshi2k2.github.io/unisub/
Our report: mp.weixin.qq.com/s/M3BypidpK…
Taha Binhuraib 🦉 retweeted
New work!
What if we used sparse autoencoders to analyze data, not models—where SAE latents act as a large set of data labels 🏷️?
We find that SAEs beat baselines on 4 data analysis tasks and uncover surprising, qualitative insights about models (e.g. Grok-4, OpenAI) from data.
Taha Binhuraib 🦉 retweeted
A left frontal-temporal network selectively supports language comprehension and production. Are computations in this language network driven primarily by bottom-up input, or by top-down task demands?
🧵👇
biorxiv.org/content/10.64898…
Two chances to catch our poster in Room 10 (upper level): 10:40–11:30 and 4:15–5:00. Stop by!
#NeurIPS2025
Taha Binhuraib 🦉 retweeted
🧠Excited to share MOSAIC: a unified fMRI mega-dataset: 430k+ trials, 160k stimuli, 93 subjects (and more to come!) all in a single brain space
Enables massive scale for modeling the visual cortex with just 2 lines of code
github.com/murtylab/mosaic-d…
🐍pip install mosaic-dataset
Headed to NeurIPS in San Diego Dec 2-7!
Always excited to dive into conversations at the intersection of neuroscience and machine learning. If you're around and want to chat about brains, machines, or where they meet, let's connect.
Looking forward to the week ahead!
🚨 Paper alert:
To appear in the DBM Neurips Workshop
LITcoder: A General-Purpose Library for Building and Comparing Encoding Models
📄 arxiv: arxiv.org/abs/2509.09152v1
🔗 project: litcoder-brain.github.io/
Taha Binhuraib 🦉 retweeted
A fun collaborative project! We leverage TunedLens (~linear decoding of tokens) to explore how LLMs' internal representations change from layer to layer.
1/
🧠 New preprint:
How Do LLMs Use Their Depth?
We uncover a “Guess-then-Refine” mechanism across layers - early layers predict high-frequency tokens as guesses; later layers refine them as context builds
Paper - arxiv.org/abs/2510.18871
@neuranna @GopalaSpeech @berkeley_ai
Taha Binhuraib 🦉 retweeted
It's been more than a year, but the EWoK (Elements of World Knowledge) paper is finally out in TACL!
tl;dr: language models learn basic social concepts way easier than physical and spatial concepts.
direct.mit.edu/tacl/article/…
💡New work!
Do LLMs learn foundational concepts required to build world models? We address this question with 🌐🐨EWoK (Elements of World Knowledge)🐨🌐, a flexible cognition-inspired framework to test knowledge across physical and social domains
ewok-core.github.io/
🧵👇
Taha Binhuraib 🦉 retweeted
As our lab started to build encoding 🧠 models, we were trying to figure out best practices in the field. So @NeuroTaha built a library to easily compare design choices & model features across datasets!
We hope it will be useful to the community & plan to keep expanding it!
1/
🚨 Paper alert:
To appear in the DBM Neurips Workshop
LITcoder: A General-Purpose Library for Building and Comparing Encoding Models
📄 arxiv: arxiv.org/abs/2509.09152v1
🔗 project: litcoder-brain.github.io/
🚨 Paper alert:
To appear in the DBM Neurips Workshop
LITcoder: A General-Purpose Library for Building and Comparing Encoding Models
📄 arxiv: arxiv.org/abs/2509.09152v1
🔗 project: litcoder-brain.github.io/
