Weather forecasts lose reliability after ~10 days, but farmers, resource managers, and communities make critical decisions weeks in advance.
New in @Nature: a lightweight ML method that corrects predictable errors in physics-based and AI weather forecasts at extended lead times.
From atoms to new materials: Yuanqi Du (Microsoft Research) introduces ATLAS & JANUS, neural samplers for metallic glasses and alloys, with applications to free energies, phase diagrams and materials discovery.
Online Monte Carlo Seminar
Nov 12 | 11:30 am ET / 4:30 pm London
Starting this spring, we have been working on how agents can actually build computational workflows that outperform human, now we are here with strong wet lab results!
The physical world is exponentially more scalable than the in silico methods that try to recapitulate it.
Come learn about @jura_bio: we compute in chemistry because the natural substrate of biological machine intelligence has never been silicon.
Our JupyterHealth team won a $10mil @LaudeInstitute Moonshot Prize! We will build an open-source health data infrastructure to bring AI-powered healthcare into people's every day lives (aka "Edge Medicine") 🚀
I'll be hiring 1-2 PhD students this cycle, so please apply 🤩
This paper will appear in the Annals of Statistics!
Personally, I quite like this work because (a) it came out of a great collaboration and (b) the core idea is simple enough to explain even to a high schooler (local linearisation).
MiniMMD 🎓 — My PhD thesis is now on arXiv!
It brings together several years of research on minimization of maximum mean discrepancy (MMD), from theory to practice.
Excited to finally share this journey and its results!
Delighted to share that @_kreda_, Yuji Nakatsukasa, and I have written a survey paper on randomized linear algebra. We tried to fill each page with our excitement for all that our research community has accomplished. Take a look! https://t.co/owf8orOC0W
To colleagues who are attending @Materials_MRS fall at Boston, the registration for the AI for Materials & Chemistry workshop at our @MSRNE office is almost full, sign-up ASAP to secure a slot and for us to consider expanding the space!
Bring your poster/talk around to engage with the community!
Can AI agents solve complex problems more efficiently? Plus, tackling coding bottlenecks; evolving agent skills from telemetry; how AI needs vary across industries; an AI tool saving teachers time; a weather forecast breakthrough; BioEmu2 in public preview. https://t.co/rU6SssWLXh
We introduce action learning: learning a stochastic action that scores how plausible an entire trajectory is, directly from observed trajectories.
No known forces. No potential energy function. No simulator likelihood.
We think this is a new paradigm for reasoning about trajectory data.
Joint work with @du_yilun and @YuanqiD.
🔥 Recovers underlying physical dynamics with over 2,000× lower error than local regression in a double well system.
🪐 Identifies a hidden 0.01% perturbation in the Sun's mass from trajectories of a 31 body orbital system.
🧬 Reuses a single learned molecular action for trajectory completion, steering, constrained optimization and temperature transfer.
The core idea: One learned action → reuse for many possible downstream tasks!
If you'd like to come work with me and my amazing colleagues as a full-time researcher @MSFTResearch@MSRNE, please apply here:
https://t.co/rXXaQuxbs6
https://t.co/F7BKiShjad
We’re hiring Senior and Principal researchers in ML & Stats at Microsoft Research New England @MSRNE! Amazing colleagues, remarkable breadth and depth of expertise, and unbeatable views. Come join us! Questions about the roles or the group? DM me!