Also, my Parisian apartment (17e arrondissement) is the cutest thing ever 😍.

Happy to be back in Paris for a while. La météo n’est pas idéale, mais le printemps est tout de même là 🌸🌼🌺. 🙂

And the codebase and open weights are now up and running! Our model DeBias-CLIP (CVPR 2026) is currently the new SOTA (by a large margin) for long text retrieval! Try it out if you’re doing image <-> text retrieval (text sequence up to 248 tokens). 🙂

We've been investing heavily in better protein language models (PLMs), but relatively little work addresses how to best generate with them. We present a new search-based method for PLMs and exhaustively benchmark models and methods, including with in vitro data from antibody therapeutics campaigns.🧵

I knew that the Department of War shot a missile into a school & killed 175 civilians, mostly children. This is learned today: Secretary of War Hegseth had cut the teams assigned to prevent such tragedies by 90%. www.npr.org/2026/03/11/n...

Anyone else tired of Linkedin AI bros clickbaits? Can we bring back good old Linkedin when real experts share useful stuff? 😔

Our freshly accepted CVPR 2026 paper is up on arxiv 😍. Project page, open code and more detailed post coming soon! « CLIP is Shortsighted: Paying Attention Beyond the First Sentence ». #CVPR

Positive vibes only. #pokopia

😂😂😂

Thank you Nintendo for always delivering at the right time. Animal Crossing for the pandemic, Pokopia for WW3 😭.

This is very cool. I kinda wish one of the big conferences would make a non-LLM submission policy for the main track just for one year. Would be fun 😛.

Project page now available! DeBias-CLIP is now the new SOTA for image-to-long-caption retrieval! trailab.github.io/DeBias-CLIP-...

😂😂😂😭😭😭

Our freshly accepted CVPR 2026 paper is up on arxiv 😍. Project page, open code and more detailed post coming soon! « CLIP is Shortsighted: Paying Attention Beyond the First Sentence ». #CVPR

In this amazing multidisciplinary collaboration, we report our early experience with the @openclaw-x.bsky.social ->

What is the relationship between memorization and generalization in AI? Is there a fundamental tradeoff? In infinitefaculty.substack.com/p/memorizati... I’ve reviewed some of the evolving perspectives on memorization & generalization in machine learning, from classic perspectives through LLMs.

Inspiring and on point quote for 2026 🥲.

BOTW ~ TOTK ~ Skyward Sword 💚💚💚

He’s so freaking cringe 😬. Is this real life? 🤯🫠

This was legit the first time I was feeling confident about a big AI conf. paper! Initial reviews were already 3/3 accepts, and we managed to get final scores of 5/5/6! Teaser: we’re now the new SOTA (by a huge margin) for image-to-long-text CLIP-style retrieval 👀.

Accepted at CVPR! More info to come soon (we need to work on our project page and all 😅). 👀

Me refreshing OpenReview every 10 seconds for CVPR decisions 👀.

😩😩😩

5 days to go until final CVPR decisions 👀👀👀. #cvpr

In terms of what cures are being lost: - Epstein-Barr virus is perhaps the major trigger for multiple sclerosis - herpes simplex virus causes cold sores, genital herpes, infections in babies, deadly meningitis - shingles virus causes an intensely painful disease

It took me weeks, but finally it's there: an overlong blogpost on synthetic pretraining. vintagedata.org/blog/posts/s...

😨😨😨

The largest Trump superPAC donor so far this cycle is the president of OpenAI

Geometry and symmetry obsession ft Berliner Dom 🇩🇪.

Coming soon #ECCV2026, Malmo 🇸🇪

Been pretty heads-down finishing Chapter 6 on implementing RLVR via GRPO. Just finished, and it might be my favorite chapter so far. Code notebook: github.com/rasbt/reason... (And it should be added to the early access soon.) The next chapter adds stability and performance improvements to GRPO.

Nano Banana Pro surprisingly very good at meme generation with perfect text 🫢. #aimeme #airesearch

Here's my enormous round-up of everything we learned about LLMs in 2025 - the third in my annual series of reviews of the past twelve months https://simonwillison.net/2025/Dec/31/the-year-in-llms/ This year it's divided into 26 sections! This is the table of contents

Excited about our new paper that unifies discrete, Gaussian, and simplicial diffusion, enabling model comparison, likelihood evaluation, stable training, and more, including a DNA design application! Amazing work from @alannawzadamin.bsky.social, Alina, Lily, and team! arxiv.org/abs/2512.15923

I have not read the paper, but I like the title, so I'm booking this for myself. "If generative AI is the answer, what is the question?" arxiv.org/abs/2509.06120

Omg that’s so so so cool! 🤩🤩🤩

Why don’t I get normal Black Friday deals in my email anymore? 😅😭😂

The claim "A does not cause B" is not an evidence-based claim because studies have not ruled out the possibility that A causes B. 😭😭😭

Dog therapy at work 🐶🐶🐶

Cool & useful survey on methodology and best practices for LLM-as-a-judge frameworks. arxiv.org/abs/2411.15594

A new essay on the crazy, all or nothing approach to work happening in AI today, the looming human costs, and the lack of a finish line. I wouldn't say it's okay, but I'm not sure how to fix it. www.interconnects.ai/p/burning-out

I follow tennis, or rather tennis follows me? 🫢😅 #viennaopen #atp500

Pretty October day.

Back in Wien, but during fall season this time, with 6 PM sunsets 🥲.

October perfectly matches the energy of witchy Prague 🎃.

Rainy witchy Praha. 🇨🇿🇨🇿🇨🇿

Berlin sunsets.

No more FOMO for Canadian fall. 📍Berlin 🇩🇪.

Finally made it to Berlin on my 4th German visit 🫢🇩🇪.

Lübeck with some sharp angles photography. Moin moin. 🇩🇪