The Bau lab is on fire ! 😍

Interested in doing a PhD at the intersection of human and machine cognition? ✨ I'm recruiting students for Fall 2026! ✨ Topics of interest include pragmatics, metacognition, reasoning, & interpretability (in humans and AI). Check out JHU's mentoring program (due 11/15) for help with your SoP 👇

Pleased to share new work with @sflippl.bsky.social @eberleoliver.bsky.social @thomasmcgee.bsky.social & undergrad interns at Institute for Pure and Applied Mathematics, UCLA. Algorithmic Primitives and Compositional Geometry of Reasoning in Language Models www.arxiv.org/pdf/2510.15987 🧵1/n

🧠 Thrilled to share our NeuroView with Ellie Pavlick! "From Prediction to Understanding: Will AI Foundation Models Transform Brain Science?" AI foundation models are coming to neuroscience—if scaling laws hold, predictive power will be unprecedented. But is that enough? Thread 🧵👇

This is so cool. When you look at representational geometry, it seems intuitive that models are combining convex regions of "concepts", but I wouldn't have expected that this is PROVABLY true for attention or that there was such a rich theory for this kind of geometry.

🕳️🐇Into the Rabbit Hull – Part II Continuing our interpretation of DINOv2, the second part of our study concerns the *geometry of concepts* and the synthesis of our findings toward a new representational *phenomenology*: the Minkowski Representation Hypothesis

🕳️🐇 𝙄𝙣𝙩𝙤 𝙩𝙝𝙚 𝙍𝙖𝙗𝙗𝙞𝙩 𝙃𝙪𝙡𝙡 – 𝙋𝙖𝙧𝙩 𝙄 (𝑃𝑎𝑟𝑡 𝐼𝐼 𝑡𝑜𝑚𝑜𝑟𝑟𝑜𝑤) 𝗔𝗻 𝗶𝗻𝘁𝗲𝗿𝗽𝗿𝗲𝘁𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗱𝗲𝗲𝗽 𝗱𝗶𝘃𝗲 𝗶𝗻𝘁𝗼 𝗗𝗜𝗡𝗢𝘃𝟮, one of vision’s most important foundation models. And today is Part I, buckle up, we're exploring some of its most charming features. :)

Superposition has reshaped interpretability research. In our @unireps.bsky.social paper led by @andre-longon.bsky.social we show it also matters for measuring alignment! Two systems can represent the same features yet appear misaligned if those features are mixed differently across neurons.

For XAI it’s often thought explanations help (boundedly rational) user “unlock” info in features for some decision. But no one says this, they say vaguer things like “supporting trust”. We lay out some implicit assumptions that become clearer when you take a formal view here arxiv.org/abs/2506.22740

🚨Updated: "How far can we go with ImageNet for Text-to-Image generation?" TL;DR: train a text2image model from scratch on ImageNet only and beat SDXL. Paper, code, data available! Reproducible science FTW! 🧵👇 📜 arxiv.org/abs/2502.21318 💻 github.com/lucasdegeorg... 💽 huggingface.co/arijitghosh/...

Check out @mryskina.bsky.social's talk and poster at COLM on Tuesday—we present a method to identify 'semantically consistent' brain regions (responding to concepts across modalities) and show that more semantically consistent brain regions are better predicted by LLMs.

1/🚨 New preprint How do #LLMs’ inner features change as they train? Using #crosscoders + a new causal metric, we map when features appear, strengthen, or fade across checkpoints—opening a new lens on training dynamics beyond loss curves & benchmarks. #interpretability

Employing mechanistic interpretability to study how models learn, not just where they end up 2 papers find: There are phase transitions where features emerge and stay throughout learning 🤖📈🧠 alphaxiv.org/pdf/2509.17196 @amuuueller.bsky.social @abosselut.bsky.social alphaxiv.org/abs/2509.05291

Mechanistic interpretability often relies on *interventions* to study how DNNs work. Are these interventions enough to guarantee the features we find are not spurious? No!⚠️ In our new paper, we show many mech int methods implicitly rely on the linear representation hypothesis🧵

I was part of an interesting panel discussion yesterday at an ARC event. Maybe everybody knows this already, but I was quite surprised by how "general" intelligence was conceptualized in relation to human intelligence and the ARC benchmarks.

Phenomenology → principle → method. From observed phenomena in representations (conditional orthogonality) we derive a natural instantiation. And it turns out to be an old friend: Matching Pursuit! 📄 arxiv.org/abs/2506.03093 See you in San Diego, @neuripsconf.bsky.social 🎉 #interpretability

🚨Our preprint is online!🚨 www.biorxiv.org/content/10.1... How do #dopamine neurons perform the key calculations in reinforcement #learning? Read on to find out more! 🧵

Are there conceptual directions in VLMs that transcend modality? Check out our COLM oral spotlight 🔦 paper! We use SAEs to analyze the multimodality of linear concepts in VLMs with @chloesu07.bsky.social, @thomasfel.bsky.social, @shamkakade.bsky.social and Stephanie Gil arxiv.org/abs/2504.11695

Check out our COLM 2025 (oral) 🎤 SAEs reveal that VLM embedding spaces aren’t just "image vs. text" cones. They contain stable conceptual directions, some forming surprising bridges across modalities. arxiv.org/abs/2504.11695 Demo 👉 vlm-concept-visualization.com

Excited to announce the first workshop on CogInterp: Interpreting Cognition in Deep Learning Models @ NeurIPS 2025! 📣 How can we interpret the algorithms and representations underlying complex behavior in deep learning models? 🌐 coginterp.github.io/neurips2025/ 1/4

How do language models generalize from information they learn in-context vs. via finetuning? In arxiv.org/abs/2505.00661 we show that in-context learning can generalize more flexibly, illustrating key differences in the inductive biases of these modes of learning — and ways to improve finetuning. 1/

Our work finding universal concepts in vision models is accepted at #ICML2025!!! My first major conference paper with my wonderful collaborators and friends @matthewkowal.bsky.social @thomasfel.bsky.social @Julian_Forsyth @csprofkgd.bsky.social Working with y'all is the best 🥹 Preprint ⬇️!!

Accepted at #ICML2025! Check out the preprint. HUGE shoutout to Harry (1st PhD paper, in 1st year), Julian (1st ever, done as an undergrad), Thomas and Matt! @hthasarathan.bsky.social @thomasfel.bsky.social @matthewkowal.bsky.social

<proud advisor> Hot off the arXiv! 🦬 "Appa: Bending Weather Dynamics with Latent Diffusion Models for Global Data Assimilation" 🌍 Appa is our novel 1.5B-parameter probabilistic weather model that unifies reanalysis, filtering, and forecasting in a single framework. A thread 🧵

Have you thought that in computer memory model weights are given in terms of discrete values in any case. Thus, why not do probabilistic inference on the discrete (quantized) parameters. @trappmartin.bsky.social is presenting our work at #AABI2025 today. [1/3]

New in the Deeper Learning blog: Kempner researchers show how VLMs speak the same semantic language across images and text. bit.ly/KempnerVLM by @isabelpapad.bsky.social ,Chloe Huangyuan Su, @thomasfel.bsky.social, Stephanie Gil, and @shamkakade.bsky.social #AI #ML #VLMs #SAEs

Firing rates in visual cortex show representational drift, while temporal spike sequences remain stable www.sciencedirect.com/science/arti... Great work by Boris Sotomayor and with @battaglialab.bsky.social

PINEAPPLE, LIGHT, HAPPY, AVALANCHE, BURDEN Some of these words are consistently remembered better than others. Why is that? In our paper, just published in J. Exp. Psychol., we provide a simple Bayesian account and show that it explains >80% of variance in word memorability: tinyurl.com/yf3md5aj

📽️Recordings from our @cosynemeeting.bsky.social #COSYNE2025 workshop on “Agent-Based Models in Neuroscience: Complex Planning, Embodiment, and Beyond" are now online: neuro-agent-models.github.io 🧠🤖

[...] overall, we argue an SAE does not just reveal concepts—it determines what can be seen at all." We propose to examine how constraints on SAE impose dual assumptions on the data, led by the amazing @sumedh-hindupur.bsky.social 😎

New paper–accepted as *spotlight* at #ICLR2025! 🧵👇 We show a competition dynamic between several algorithms splits a toy model’s ICL abilities into four broad phases of train/test settings! This means ICL is akin to a mixture of different algorithms, not a monolithic ability.

Want strong SSL, but not the complexity of DINOv2? CAPI: Cluster and Predict Latents Patches for Improved Masked Image Modeling.

🚨 New Paper! Can neuroscience localizers uncover brain-like functional specializations in LLMs? 🧠🤖 Yes! We analyzed 18 LLMs and found units mirroring the brain's language, theory of mind, and multiple demand networks! w/ @gretatuckute.bsky.social, @abosselut.bsky.social, @mschrimpf.bsky.social 🧵👇

Scaling Laws in Patchification: An Image Is Worth 50,176 Tokens And More Feng Wang, Yaodong Yu, Guoyizhe Wei, Wei Shao, Yuyin Zhou, Alan Yuille, Cihang Xie tl;dr: we trained 1px patch size ViT so you don't have to. It improves results, but costly. arxiv.org/abs/2502.03738

I'm delighted to share this latest research, led by the talented @hthasarathan.bsky.social and Julian. Their work uncovered both universal conceptual across models but also unique concepts specific to DINOv2 and SigLip! 🔥

🌌🛰️🔭Wanna know which features are universal vs unique in your models and how to find them? Excited to share our preprint: "Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment"! arxiv.org/abs/2502.03714 (1/9)

Using mechanistic #interpretability 💻 to advance scientific #discovery 🧪 & capture striking biology? 🧬 Come see @jhartford.bsky.social's oral presentation 👨‍🏫 @ #NeurIPS2024 Interpretable AI workshop 🦾 to learn more about extracting features from large 🔬 MAEs! Paper 📄 ➡️: openreview.net/forum?id=jYl...

Did you know that @PyTorch implements the Bessel's correction to standard deviation, but not numpy or jax. A possible source of disagreements when porting models to pytorch! @numpy_team

New preprint: "The geometry of the neural state space of decisions", work by Mauro Monsalve-Mercado, buff.ly/42wVHD5. Surprising results & predictions! (Thread) We analyze neuropixel population recordings in macaque area LIP during a reaction time, random-dot motion 1/

How do tokens evolve as they are processed by a deep Transformer? With José A. Carrillo, @gabrielpeyre.bsky.social and @pierreablin.bsky.social, we tackle this in our new preprint: A Unified Perspective on the Dynamics of Deep Transformers arxiv.org/abs/2501.18322 ML and PDE lovers, check it out!

The research introduces AXBENCH, demonstrating that prompting outperforms complex representation methods like sparse autoencoders. It features a new weakly-supervised method, ReFT-r1, which combines interpretability with competitive performance. https://arxiv.org/abs/2501.17148

The application for our #KempnerInstitute graduate fellowship for Harvard Ph.D. students is now open! Register for an upcoming virtual open house (2/6 & 2/18) and apply by 3/1. Details: bit.ly/40RY6qS

DinoV2, C:5232... 😶‍🌫️

Does the culture you grow up in shape the way you see the world? In a new Psych Review paper, @chazfirestone.bsky.social & I tackle this centuries-old question using the Müller-Lyer illusion as a case study. Come think through one of history's mysteries with us🧵(1/13):

Join us at our upcoming workshop at ICLR, XAI4Science: From Understanding Model Behavior to Discovering New Scientific Knwlg, Apr 27-28 submissions on a-priori (ante-hoc) & a-posteriori (post-hoc) interpretability & self-explainable models for understanding model’s behvr welcm tinyurl.com/3w8sddpm

The Mathematics of Artificial Intelligence: In this introductory and highly subjective survey, aimed at a general mathematical audience, I showcase some key theoretical concepts underlying recent advancements in machine learning. arxiv.org/abs/2501.10465

New paper! A self-supervised object-centric 2.1D image representation using 3D Gaussians, extending MAE with a Gaussian bottleneck. While Gaussian splatting has been used for single-scene reconstruction, we’re the first to apply it to image representation learning! brjathu.github.io/gmae/.

New paper <3 Interested in inference-time scaling? In-context Learning? Mech Interp? LMs can solve novel in-context tasks, with sufficient examples (longer contexts). Why? Bc they dynamically form *in-context representations*! 1/N

*Understanding Visual Feature Reliance through the Lens of Complexity* by @thomasfel.bsky.social @louisbethune.bsky.social @lampinen.bsky.social Wonderful work! They rank features' complexity with a variant of mutual information, before analyzing their dynamics. arxiv.org/abs/2407.06076