Micha Heilbron

@m_heilb

Group leader @MPI_NL (LPC group) // Assistant Professor of Cognitive AI @UvA_Amsterdam, @AmsterdamBrain. Cog-sci 🤝 AI 🤝 neurosci. 🟦 mheilbron.bsky

Amsterdam, NL

lpc-lab.github.io

Joined June 2015

1,420 Following

2,126 Followers

Replies Reposts Media

Micha Heilbron
@m_heilb
Oct 26
In Budapest for #EMNLP2026. Two papers today! 11:00, Poster Session D: @KarolinaADrozdz on entity tracking in language models vs humans 12:18, talk in Psycholinguistics 2: @dr_han_jones on human-like fleeting memory in transformers Will be around all week -- come say hi!
Karolina Drożdż
1
Micha Heilbron
@m_heilb
Oct 10
New, strongest evidence yet against the "Octopus Argument" by Bender & Koller A language model trained on text alone & a vision model trained just on images learn *such* similar representations that you can roughly match images to captions without any paired example great paper
Phillip Isola
@phillip_isola
Oct 9
This is a result I’ve dreamt about for many years: *unpaired translation between images and text* I thought it might be only slightly possible, the kind of thing you have to really squint at. But Dominik proved this wrong. You don’t have to squint. Worth looking for yourself:
Quote
Dominik Schnaus
@dominik_schnaus
Oct 9
DINOv2 has never seen a caption, and Qwen3 has never seen an image. We still aligned their embedding spaces without a single image-caption pair. It even works when the images and the captions come from different datasets. Project page: dominik-schnaus.github.io/unpaired-roset… ⬇️
2
Micha Heilbron
@m_heilb
Sep 2
Re: the moral panic about anthropomorphism in the coverage of the OpenAI incident Talking about what an AI agent "wanted" or "tried" isn't sloppy, irresponsible anthropomorphism, but can be a perfectly legitimate way to explain its behaviour This thread lays out why (and when)
Raphaël Millière
@raphaelmilliere
Sep 1
I see @dwarkesh_sp's piece about the recent OpenAI/Huggingface incident reignited endless debates about the dangers of anthropomorphism and the legitimacy of intentional glosses of AI agent behavior, so here's a philosophical perspective on this. 1/22 x.com/dwarkesh_sp/st…
1
Micha Heilbron
@m_heilb
Aug 21
Now accepted at #EMNLP Main! 🇭🇺 Huge congrats to @KarolinaADrozdz -- first-authored and accepted before even starting her PhD
Micha Heilbron
@m_heilb
Aug 20
New preprint, w/ @KarolinaADrozdz Understanding language requires internally modelling what's happening where, even when unstated We built a new task to measure this ability in language models and people, and find it emerges at model scales far smaller than previously thought
1
Micha Heilbron
@m_heilb
Aug 20
New preprint, w/ @KarolinaADrozdz Understanding language requires internally modelling what's happening where, even when unstated We built a new task to measure this ability in language models and people, and find it emerges at model scales far smaller than previously thought
5

Micha Heilbron

Micha Heilbron’s posts