Bruce (Zhi) Wen

Why I took a 4.5 hr $90 flixbus to Ottawa instead of a 4.5 hr $500 train when Air Canada messed up ⬇️

With fresh support of $75M from NSF and $77M from NVIDIA, we’re set to scale our open model ecosystem, bolster the infrastructure behind it, and fast‑track reproducible AI research to unlock the next wave of scientific discovery. 💡

I’m surprised! Maybe the Reddit communities aren’t good comparisons (they have their own norms)? Or the random seed words are more random than prior work? Or it just really helps to prompt with the start of a story?

As co-organizer, Mila is excited to invite you to the AIMS Hackathon 2025! This global online event will unite developers, researchers, and human rights advocates to create innovative AI solutions to combat modern slavery.
📅 Registration open until August 20 fundacionpasoslibres.org/aimshackathon/

My least popular (and most correct) view is that cars should be automatically limited to the local speed limit. Put the pedal to the floor and you still can't go over 25mph in a residential area.

(15 in Manhattan btw)

I pray to god this project will pan out without too much delay and being watered down, and I’m not even religious. Meanwhile I’ll keep contemplating moving to Europe.

I feel the same way traveling between Toronto and Montreal too. Worst part is the existing train is slower AND more expensive than flying (and has the same luggage restriction).

EPFL NLP Postdoctoral Scholar Posting - Swiss AI LLMs The EPFL Natural Language Processing (NLP) lab is looking to hire a postdoctoral researcher candidate in the area of multilingual LLM design, training, and evaluation. This postdoctoral position is as...

The EPFL NLP lab is looking to hire a postdoctoral researcher on the topic of designing, training, and evaluating multilingual LLMs:

docs.google.com/document/d/1...

Come join our dynamic group in beautiful Lausanne!

That’s one way to add diversity

Monotonous yes, but I really enjoyed my samplings of Rome and Barcelona! Are UK cities good because they’re diverse?

Hopped to Salzburg for 24 hr too. Foggy and moody weather fits the hilly town perfectly. Meanwhile the Salzburg festival has amazing energy.

Some photos from Vienna 🇦🇹 #ACL2025

A pleasant, human-scale city with reminders of its imperial past everywhere. Wish it had a better food scene though, or maybe Canadian cities raised the bar too high …

Who would have thought the REALM workshop with its lineup would be popular 🤦‍♂️ #ACL2025

On this World Day Against Trafficking in Persons, discover Project AIMS (AI against Modern Slavery), a Mila initiative that leverages AI to efficiently analyze corporate reports submitted under the Modern Slavery Act (MSA): ow.ly/FXXW50WxpTG

Our new paper in #PNAS (bit.ly/4fcWfma) presents a surprising finding—when words change meaning, older speakers rapidly adopt the new usage; inter-generational differences are often minor.

w/ Michelle Yang, ‪@sivareddyg.bsky.social‬ , @msonderegger.bsky.social‬ and @dallascard.bsky.social‬👇(1/12)

The announcements at #ACL2025 sound exactly like those in Squid Games it’s uncanny.

“It was carefully submitted as by A.H. Land and A.G. Doig with no indication whatever that we were both women.”

I’ll be at ACL, hit me up if you’re interested in research scientist / senior dev positions at Mila! (Or just chat that’s fine too)

#ACL2025 #ACL2025NLP #NLProc

I’ll be at ACL, hit me up if you’re interested in research scientist / senior dev positions at Mila! (Or just chat that’s fine too)

#ACL2025 #ACL2025NLP #NLProc

In 1965, Margaret Dayhoff published the Atlas of Protein Sequence and Structure, which collated the 65 proteins whose amino acid sequences were then known.

Inspired by that Atlas, today we are releasing the Dayhoff Atlas of protein sequence data and protein language models.

XAI’s dogwater performance on the 2025 IMO confirms that their Grok 4 benchmark claims were hot air. Their eye popping metrics were down to the following innovations:
- train on test
- train on test
- train on test

My thoughts on the crucial importance of methodology on self-reported AI performance on mathematics competitions, and my policy on commenting on such reports going forward: mathstodon.xyz/@tao/1148814...

Such a useful framing (which I first heard from @alexanderhoyle.bsky.social). The prompt is ultimately just a bunch of numbers that we input to the model, and why we should prefer one set of numbers over another set is hard to answer, as we can’t search the space the way we normally would in ML.

The #ACL2025 #ACL2025NLP feed is up and running! It matches both hashtags and any posts from or mentions of @aclmeeting.bsky.social

Pin it to your home 📌 and enjoy!

bsky.app/profile/did:...

One of the things that the creator enjoyed more than customers …

If you 1) want to work with cutting-edge AI research, 2) want to have real-world impact, and 3) don’t care about chasing profits, then this might be your dream job.

Développeur.se Principal.e - Recherche Appliquée en Apprentissage Automatique - Mila - Institut québécois d'intelligence artificielle À propos de MilaFondé par le professeur Yoshua Bengio de l’Université de Montréal, Mila rassemble des chercheurs spécialisés en intelligence artificielle et plus précisément en apprentissage automatiq...

We’re hiring senior devs and applied research scientists!

(yes Mila has a dedicated applied research team, and yes it’s a great place to work)

apply.workable.com/mila-2/j/0EF...

apply.workable.com/mila-2/j/1C3...

A very interesting read: goombalab.github.io/blog/2025/tr...

Tokenization was one of my least favourite parts of the vanilla Transformer LMs (the other was additive positional encoding). There's no inherent reason why tokenization is needed for language and I hope we can move past it.

"in 2025 we will have flying cars" 😂😂😂

These are good tips. One thing I want to add is, if I can’t put thoughts into words precisely, I’d first write them down in the crudest, roughest, least polished way possible as long as the ideas are clear. Then I’d refine the text, use fancier words, connect better, make more concise, etc.