Ron Shprints

Recordings have just dropped on YouTube! You should definitely check out this amazing resource if you want to learn about flow matching. Peter is the best of the best and he put so much effort into this. It was a pleasure to contribute my small part:)

@andrewgwils I started as a chemist. Took a long detour in "pure" AI for a while. Now I partake in the so called "AI4Science" party. Am I a scientist? Or an ML researcher? I wouldn't know. I'm probably neither. But I certainly enjoy both.

Presently, many ML researchers dabble in scientific applications, as part of what is called "AI for Science". In the future, domain scientists will become AI scientists, and what it means to be an AI researcher will evolve to be more narrowly focused on foundations.

@andrewgwils I'm going to be nerdsniped for this, but when I figured this in high school I was absolutely blown away. For weeks, my go to conversation starter was the microwave story. I just couldn't believe that we managed to put this machine together. Oh god I've just got nerdsniped again.

The microwave must have been mindblowing when it came out. Even now, it's so much more futuristic and cool than "cutting edge" AI technology. The most sci-fi device in our lives.

Most people assume you need a massive dataset to distill flow models. We challenge that. Is data actually necessary? Or perhaps it is a liability? Introducing FreeFlow: We achieve SOTA (1.49 FID on ImageNet-512) 1-step image generation without a single data sample. 🧵👇[1/n]

@owl_posting That sounds like a paper that I'd write

"Analyzing NBA player positions and interactions with density-functional fluctuation theory" nature.com/articles/s4159…

truly the greatest day ever🎗️

New work led by @peholderrieth showing how to transform an already trained flow matching model to a stochastic transition/posterior model that can still be sampled via an efficient ODE solver!

New work: “GLASS Flows: Transition Sampling for Alignment of Flow and Diffusion Models”. GLASS generates images by sampling stochastic Markov transitions with ODEs - allowing us to boost text-image alignment for large-scale models at inference time! arxiv.org/pdf/2509.25170 [1/7]

@TrevorCampbell_ @AllThingsApx Nah it's definitely the Kendall Square Cava

@AllThingsApx Kendall square Tatte and Area Four

What is the AI x Bio equivalent of the Pentagon Pizza Report?

@y0b1byte How are you using cursor with your setup?

Started using cursor:

a few months ago we @_DimensionCap quietly funded a team of 3 MIT computer scientists & mathematicians at the forefront of ML & physics+simulation for enzyme design. these. guys. are. cracked today they're stealthily building the early team. if you're world class, look below 👇

@GeminiApp @Noahpinion

Trampolines aren't the only things bunnies are into #veo3

@dvruette Wdym by poor shuffling in the data loader? Ordering is similar to the original dataset?

issues were numerical stability of loss function (fixed by manual mixed precision) and poor shuffling in the streaming data loader (fixed by dask.dataframe -> df.sample(frac=1.0))

at last, it is fixed! there shall be no more funny-looking loss curves from here on out 🙏

Thanks @MIT_CSAIL for featuring our work!🖊️🎨 Huge thanks to the CSAIL news team for the fun article + video!! We'll be presenting SketchAgent at #CVPR2025 next week — come say hi if you're curious how LLMs can be used to collaboratively sketch!🖌️ 👉 bit.ly/43mTme1

@skalskip92 Super cool! Have you tried to use it to collect statistics? For example, distance traveled, speed, etc. Can you generalize it to do offside detection? Perhaps with uncertainty quantification too?

here’s the code if you want to play around: github.com/roboflow/noteb…

"can you do the same thing for football?" - I got this question 100x today yep, I actually did this last year - player detection and tracking - team clustering - perspective transformation end-to-end YT tutorial: youtube.com/watch?v=aBVGKo…

Some personal updates: 1. I will defend my PhD thesis, "Advancing Artificial Intelligence for Efficient and Synthesizable In-silico Molecular Design," at 10 am on May 9th in MIT 66-110, all are welcome! 2. After a ~1 year postdoc at Stanford, I’ll join the CBE department at UPenn as an AP to continue my research on systematizing molecular discovery methodologies by developing chemically and physically plausible AI methods. Feel free to reach out if you have a shared interest!

@mmbronstein Ready to chart the manifold’s course.

A short ride in beautiful Oxfordshire before GDL class

@karpathy LOL

There's a new kind of coding I call "vibe coding", where you fully give in to the vibes, embrace exponentials, and forget that the code even exists. It's possible because the LLMs (e.g. Cursor Composer w Sonnet) are getting too good. Also I just talk to Composer with SuperWhisper so I barely even touch the keyboard. I ask for the dumbest things like "decrease the padding on the sidebar by half" because I'm too lazy to find it. I "Accept All" always, I don't read the diffs anymore. When I get error messages I just copy paste them in with no comment, usually that fixes it. The code grows beyond my usual comprehension, I'd have to really read through it for a while. Sometimes the LLMs can't fix a bug so I just work around it or ask for random changes until it goes away. It's not too bad for throwaway weekend projects, but still quite amusing. I'm building a project or webapp, but it's not really coding - I just see stuff, say stuff, run stuff, and copy paste stuff, and it mostly works.

Machine learning is a huge field today, but at its core it's about two principles: 1) Generalization 2) Scaling laws

To be clear, this is not criticism, I'm genuinely curious

I watched Barca vs. Betis today and this offside call by VAR led to the cancelation of a goal. I'm not familiar with the semi-automatic offside system, but I wonder whether current technology is accurate enough to make such a call. Why are there no uncertainty bars in this pic?