
When you say, "This is a reinforcement learning problem," you should say it with the same excitement as "This is NP-hard."

machine learning, reinforcement learning, programming languages, handstands (he/him)

When you say, "This is a reinforcement learning problem," you should say it with the same excitement as "This is NP-hard."


How do you test a randomized algorithm? Running it a million times might not be enough. I wrote about a few testing strategies, culminating in a neat trick: enumerate every possible execution trace and check correctness exactly. timvieira.github.io/testing-…

Conditional Poisson Sampling (CPS) CPS is a distribution over fixed-size subsets in which each subset’s probability is proportional to the product of its items’ weights. This one has lots of fun JavaScript and math. timvieira.github.io/conditio…

Late to the party, but this blog post is excellent. Arguably the way people should first learn neural networks.
how much can models remember? there’s a neat perspective on neural networks that deserves to be more widely known here’s how you could have invented every layer of the transformer from memory (1/n) from my new blog post!

The Token Game (srush.github.io/the-token-ga…) It's BPE the mobile game? No reason for this to exist. I'm absolutely terrible at it.

I've been thinking about where the data for the next generation of open models will come from. The internet gave us a remarkable commons for pretraining. Posttraining is different: it consumes tasks, not documents. And that commons doesn't exist. Not yet. openathena.ai/blog/there-is-…

My dream had always been to prove beautiful theorems as the AI agents unloaded my dishwasher. Well, it went the wrong way... 😂

🚢Marin 535B-A23B training has crossed the halfway mark! Some observations and learnings so far:

Sampling to Reinforce (srush.github.io/sampling-to-…) An elementary tutorial on sampling w/o randomness with a focus on gaining intuition for variance reduction as the key challenge.

Lean Verified Transformers (srush.github.io/lean-transfo…) In which we prove a bunch of Transformer invariants from scratch in Lean, and speculate about how hard it would be to do that for the rest of the world's code.

Hot take: I think it's still important to understand the code that our agents write! In this mega thread (based on my AIE talk today), I will explain why that's the case, and show some ideas for how to efficiently understand code. Alright, let's dive in. 1/

Slides from my ICML tutorial "Is numerical optimization theory irrelevant to machine learning practice in 2026?": cs.ubc.ca/~schmidtm/Document… (Updated to fix some typos, incorporate feedback, and add some things I did not have time for. Will link to the video here when posted.)

Fantastic slides from the ICML tutorial of @MarkSchmidtUBC (“is opt theory relevant in 2026”) cs.ubc.ca/~schmidtm/Document…

@ReviewAcl Why does every author have to fill out the registration form if they already have an OpenReview profile?

It’s deployment time! You’ve done the pre-deployment evals. You THINK your model is safe, so you ship it 🚀 🚨 After deployment, reports of misbehavior start trickling in What happened?? How could you have caught it?? 🤔 @icmlconf 2026 Spotlight! 🧵