(1/8) Happy to share that our paper LieFlow has been accepted to #ICML2026.

Symmetry is a strong inductive bias in ML, but you have to know the symmetry in advance. Can a model discover it directly from data, for both discrete and continuous groups? 🧵

How should we build long-horizon manipulation systems? As coding agents become more capable, what interface should connect models to the physical world? We’ve been thinking about this since 2024.

Introducing Retriever, a programming framework for asynchronous, closed-loop robot past year at Stanford and Northeastern.

- Robot agents run in physical time with various feedback loops, currently lacking a good programming abstraction.
- Retriever introduces a time-aware building block, “Flow”, analogous to a PyTorch layer or a Ray Actor.
- Users connect Flows and specify when each runs and what inputs it reads. Retriever handles asynchronous execution.
- Agents are represented as directed graphs with temporal feedback loops for closed-loop, hierarchical systems, decoupling algorithm logic from execution.

We built Retriever-0 agent to demonstrate long-horizon, partially observable tasks combining perception, memory, reasoning, and manipulation.

Here's the system in action. Paper, code, and demos below.

Long horizon bimanual mobile manipulation requires reasoning in many coordinate frames: base, L/R hands, etc. In which frame would policy work best?
It really depends, so don’t pick. Mixture of Frames Policy: denoise in multiple frame in parallel.
🌐mofpo.github.io (1/9)

I'll be @icmlconf presenting our work Smoothness Errors in Dynamics Models and How to Avoid Them, with Luisa Li, @park_jungy , and @RobinSFWalters !!

Come see our poster #2200 , july 9th, 2:30 PM – 4:15 PM, hall A.

Links to paper, website, and code in thread!

Come check out symmetry discovery with flow matching today 5pm at Hall A #3010!

A couple of us @ProfluentBio will be at ICML in Seoul next week! Reach out to me if you want to talk about protein language models, equivariance, and generative models. Coffee chats and dinner signups in the post below.

Self-conditioning can be distilled into one-step language generation.

We show that self-conditioned flow LMs perform fixed-point iteration, and introduce fixed-point flow maps to compress both the iterations and the entire flow.

At one step, FMLM★ achieves 112.5 gPPL at ), substantially improving over FMLM’s 168.3 gPPL at 5.17 entropy.
FMLM★ also achieves the best 2–4 step results.

📎 Paper: arxiv.org/abs/2607.00714
⌨️ Code: github.com/Ugness/self-co…

Clever use of unitary convolutions to overcome graph oversmoothing, but relax them to tailor to the type of dynamics at hand (e.g. diffusive). Excellent work led by @bermanatorCodes and Luisa Li.

Come see us and the poster at #ICML2026 🇰🇷

(7/8) On MI-Motion (real human motion capture), no target symmetry is imposed; it must be learned from data. People move along axis-aligned directions while gravity breaks full SO(3), inducing an approximate C4 about the z-axis, which LieFlow recovers while LieGAN does not.

(8/8) Links:
🌐 jypark0.github.io/lieflow/
📄 arxiv.org/abs/2512.20043
💻 github.com/jypark0/lieflow

Joint work with @YuxuanChen23, @FEijkelboom, @jiankeyang, @jwvdm, Lawson L.S. Wong, and @RobinSFWalters.

AI has two modes in drug discovery. Accelerate: moving faster through the existing playbook. Unlock: opening frontiers that weren't possible before.

Excited to announce Profluent is partnering with Eli Lilly, the global pharma powerhouse, to unlock breakthrough medicines for eal beyond the numbers ($2.25B + royalties): we’ll get to use our frontier AI models and foundational datasets to design proteins focused on large gene insertion, a therapeutic moonshot.

Proteins govern almost everything in biology. We've built a generalizable AI platform to design all proteins. Onward!

Boston Symmetry Day is happening TODAY at Northeastern University’s Columbus Place and Alumni Center (716 Columbus Ave, 6th floor)! Breakfast starts at 9 AM, but talks are happening throughout the day, followed by a social. We’ll see you there!

Registration is now open for Boston Symmetry Day on March 31! Sign up by March 21st at docs.google.com/forms/d/e/1FAI…

We have an exciting lineup of speakers (see our website: bostonsymmetry.github.io ) 
Also featuring a poster session so you have a chance to present your awesome work!

Save the date -- Boston Symmetry Day 2025 will be held on March 31st, at Northeastern University!

Speakers and sponsors to be announced in the coming weeks, but you can expect another great lineup of talks, networking, and posters. We'll see you there!

Excited for our fall poster session and social tomorrow at MIT!

Details: The poster session will be held from 4:30 to 6pm on the first floor of MIT's Stata Center, followed by a group walk over to a nearby bar/restaurant.

See you there!