The visual world is composed of objects, and those objects are composed of features. But do VLMs exploit this compositional structure when processing multi-object scenes? In our 🆒🆕 #ICLR2026 paper, we find they do – via emergent symbolic mechanisms for visual binding. 🧵👇

📢 I am hiring a highly-motivated Ph.D student at the University of Copenhagen to work on tokenization-free NLP. Read our previous work in this topic: aclanthology.org/2025.emnlp-m... aclanthology.org/2023.emnlp-m... openreview.net/forum?id=FkS... Apply by March 8: employment.ku.dk/phd/?show=1563

nooo claude's --verbose is gone/broken, how will i now catch wrong assumptions before it goes on for 10 minutes?

A paper that should get more attention, for those interested in building truly multimodal models (aka not just plugging stuff post hoc into LLMs): arxiv.org/abs/2412.06646 TLDR: Counterintuitively, native multimodal modes seem *less* unified internally (interp)

Epstein’s economic power among academics was made possible by a capitalist system that makes higher education dependent on the charity economy rather than a public good supported by taxing the rich

🎉 Excited to share our new paper which was accepted to #AAAI2026! As LLMs become increasingly used as sources of factual knowledge, we ask: Do they perform equitably across users of different backgrounds? 🧵⬇️ 1/6

Figuring out the right rules+practices for my CLAUDE.md file and generally how to prompt Claude Code, has taught me more about software/ML engineering paradigms/concepts/best-practices than anything I've learned in uni

Listenining to Michelle Obama's audiobook "Becoming" (loving it btw) in 2026 is wild, an almost comical contrast to now... What hopeful times it was back then, as I'm now at the chapters (2006-2008) describing their 2008 run for president

Wild output from Gemini i got two weeks ago i wonder if the appearance on tv is meant to represent the sin of pride

With the latest coding agents there's almost no excuse anymore to publish a paper without any cool demo or interactive data explorer For my current paper it made it so much easier to quickly grasp different interp tools and their effects throughout the whole project

New blog post 📜 "Better late than never: Getting into interpretability in 2025" It's been a great year pivoting into interp and i wanted to reflect on it bennokrojer.com/interp.html

Really enjoyed the discussions in the UT Austin NLP group!

Couldn't have wished for a better place to do my PhD, come apply!

I'll be at COLM! Excited to chat about about anything vision+language, interpretability, cogsci/psych, embedding spaces, visual reasoning, video/world models

Devoured this book in 18 hours, usually not a big fan of audio books! It covered lots from crowdworker rights, the ideologies (doomers, EA, ...) and the silicon valley startup world to the many big egos and company-internal battles Great work by @karenhao.bsky.social

Lmao

Congratulations @bennokrojer.bsky.social on passing your PhD proposal exam! A great presentation and exciting work!

very happy to see the trend of a Behind the Scenes section catching on! transparent & honest science 👌 love the detailed montreal spots mentioned consider including such a section in your next appendix! (paper by @a-krishnan.bsky.social arxiv.org/pdf/2504.050...)

Super cool work on quantifying with NLP how language evolves through generations In linguistics, the "apparent time hypothesis" famously discusses this but never empirically tests it

I miss Edinburgh and its wonderful people already!! Thanks to @tallinzen.bsky.social and @edoardo-ponti.bsky.social for inspiring discussions during the viva! I'm now exchanging Arthur's Seat for Mont Royal to join @sivareddyg.bsky.social's wonderful lab @mila-quebec.bsky.social 🤩

Started a new podcast with @tomvergara.bsky.social ! Behind the Research of AI: We look behind the scenes, beyond the polished papers 🧐🧪 If this sounds fun, check out our first "official" episode with the awesome Gauthier Gidel from @mila-quebec.bsky.social : open.spotify.com/episode/7oTc...

Turns out condensing your research into 3min is very hard but also teaches you a lot Finally the video from Mila's speed science competition is on YouTube! From a soup of raw pixels to abstract meaning t.co/RDpu1kR7jM

Excited to share the results of my recent internship! We ask 🤔 What subtle shortcuts are VideoLLMs taking on spatio-temporal questions? And how can we instead curate shortcut-robust examples at a large-scale? We release: MVPBench Details 👇🔬

Top 99% in Boston 💪 Love these interactive maps

When you forget your leftover rice in the fridge for 3 weeks

Maybe it's just that I'm now paying more attention to the good parts again but since this post bluesky seems more fun again. Still not many paper discussions going on but saw some fun general posts

It was tough just logging back into my retired twitter account and to see a timeline that is so much fuller with interesting research discourse... And yes I've tried to customizing my feeds and whatnot but no feed can fix a lack of posts

I finish my work day with the conclusion that code assistants are maybe net negative for my short-term progress and most likely negative for my long-term progress and learning Also sycophancy is annoying af

Attend my AI 2025 bootcamp

This pattern is going to repeat in one domain after another, and gradually force us to admit that 60% of every job is networking, knowing who to trust, and doing poorly justified risk/benefit assessment.

with great responsibility comes a great amount of reimbursements to file

are there any analysis/interp papers on image editing? papers that are more insights than performance

Day 13: (the original mega-thread has become too long and nested so reposting now as a new strategy) Patchscopes: A Unifying Framework for Inspecting Hidden Representations of Language Models A few notes below 👇 I took less digital notes this time as i was sitting outside in the sun reading 🌞

This is one of the most-shared posts on Bluesky in the past day and it's just completely false. You might think ChatGPT is a *bad* search engine, or prefer another search engine. But it has had integrated web search since last year.

A must-read for anyone in NLP right now

Great work from labmates on LLMs vs humans regarding linguistic preferences: You know when a sentence kind of feels off e.g. "I met at the park the man". So in what ways do LLMs follow these human intuitions?

I would only use if it there were direct verbatim quotes from sources mentioned as the primary answer. Otherwise same, scroll past it

Super useful! I didn't actually know this trick

I will miss Thomas Müller One of the greats of football

This week, Mila researchers will present more than 90 papers at @iclr-conf.bsky.social in Singapore. Every day, we will share a schedule featuring Mila-affiliated presentations. Day 1 👇 #ICLR2025 mila.quebec/en/news/foll...

Fulfilling the male cliche and learning about the fall of the Roman empire which is definitely not related to any current events

Now the paper is also on arxiv and can be easily cited: arxiv.org/abs/2504.07128

My friend @sshkhr.bsky.social created this cool tool to track all kinds of ML summer schools - check it out!

Has anyone had a situation where an arxiv submission is on hold for more than a week? What are usually causes for this?

My labmates made a cool thing The video is also fun to watch!

Really enjoyed diving deep into this paper today: arxiv.org/abs/2406.16860 It's so systematic and a treasure trove full of insights

Here is a great summary of the paper: www.youtube.com/watch?v=aO_c... @sivareddyg.bsky.social gave a talk today at the Simon's institute workshop on The Future of Language Models and Transformers

Never been part of a project like this before - it was a very rewarding+unique experience! Everyone in the lab contributed different chapters and it was much more exploratory than your average phd project. My chapter studied R1's reasoning on "image generation/editing" (via ASCII) 🧵👇 1/N

Which evals do you pay attention the most when judging how good a new VLM is?