Michael Albergo (@msalbergo) / X

Michael Albergo

1,512 posts

Michael Albergo

@msalbergo

prof @Harvard and @KempnerInst and stuff

Joined February 2018

Pinned
Michael Albergo
@msalbergo
Apr 26
New paper! Presenting Discrete Flow Maps: paper: arxiv.org/abs/2604.09784 blog: malbergo.me/discrete-flow-… A laughable problem for me these days is that @nmboffi and I share a research brain, and we have had, time and again, a conversation that ends with “ha so I guess we’re
GIF
75K
Michael Albergo
@msalbergo
Jan 21, 2021
We’ve put out a tutorial on normalizing flows in lattice field theory! Lots of details on building equivariant and manifold flows (with code!) and some examples in scalar and gauge theory —@KyleCranmer @DaniloJRezende @sracaniere arxiv.org/abs/2101.08176
Michael Albergo
@msalbergo
Feb 2, 2023
Our paper on a general framework for efficiently building continuous normalizing flows between any distributions has been accepted @ICLR 2023! Here are some flow-flowers. of interest: @DaniloJRezende @KyleCranmer @FrankNoeBerlin @ylecun @wgrathwohl Paper: openreview.net/forum?id=li7qe…
GIF
52K
Michael Albergo
@msalbergo
Jul 25, 2024
🌞 One final bit of news — excited to announce I’ll be starting as an assistant professor in applied mathematics @Harvard and @KempnerInst Investigator in 2026 :) Please reach out if you may be interested in working with me. Grateful for everyone that made this possible!
Kempner Institute at Harvard University
@KempnerInst
Jul 25, 2024
We are thrilled to announce the appointments of three new #KempnerInstitute Investigators who will play key roles in shaping & advancing our research program. Learn more about @xkianteb, @du_yilun and @msalbergo in our official announcement: bit.ly/3WFoG4E @hseas #AI #ML
33K
Michael Albergo
@msalbergo
Jan 11, 2024
I finished my PhD :) thanks to everybody who supported me and whom I had the privilege to learn from. And thanks endlessly to @KyleCranmer who has been an incredibly supportive, advisor, collaborator, and friend.
Kyle Cranmer
@KyleCranmer
Jan 11, 2024
Congratulations to @msalbergo on successfully defending his PhD thesis! It's been an honor and a pleasure to advise you on your journey. Very impressive multidisciplinary work, check it out: scholar.google.com/citations?user…
44K
Michael Albergo
@msalbergo
Apr 2, 2024
Life update! In the fall I'll start as a Junior Fellow at the Society of Fellows at @Harvard and an @iaifi_news fellow at MIT. Looking forward to new adventures with folks across @KempnerInst @hseas @MITMath and @harvardphysics. Please reach out if/when you're in Cambridge 🙂.
30K
Michael Albergo
@msalbergo
Oct 11, 2023
🥳New paper thread🥳: We show how to build flows/diffusions between *dependent distributions*. The benefits seem really cool! Stochastic Interpolants with data-dependent couplings: arxiv.org/abs/2310.03725, big thanks @marikgoldstein @nmboffi Rajesh Ranganath, Eric Vanden-Eijnden
GIF
38K
Michael Albergo
@msalbergo
Dec 30, 2020
Marsaglia's algorithm allows one to uniformly sample from an (n-1)-sphere by rescaling normal deviates. You can easily do it in @PyTorch!
Michael Albergo
@msalbergo
Jun 14, 2024
📜💡New paper with the wonderful @nmboffi and @evdende2! We show a variety of learning paradigms for the two-time flow map associated to a probability flow. arxiv.org/abs/2406.07507 This allows us to define a handful of approaches for creating a 1-to-few step generative model:
13K
Michael Albergo
@msalbergo
Dec 23, 2020
Flows in ML pop up all over the place. There's lots of choices + research directions that go into how we choose to transport data under a target density or with MLE. It can be fun to visualize these maps to get a sense of how they differ, and at what costs + benefits!
Michael Albergo
@msalbergo
Oct 6, 2020
Grateful for having met Penrose at ⁦@Perimeter⁩ fireside chats last year — amazing how his work appears everywhere you look
Michael Albergo
@msalbergo
Jun 11, 2021
*New paper* with the @MIT crew, @DaniloJRezende @sracaniere @DeepMind , @KyleCranmer and Julian Urban! We construct normalizing flows that are compatible with sampling path integrals of quantum field theories involving fermions. Fermions make this tricky! arxiv.org/abs/2106.05934
Michael Albergo
@msalbergo
Jan 4, 2024
One week before final thesis deadline and my life looks like: h/t @KyleCranmer
7.8K
Michael Albergo
@msalbergo
Sep 15, 2020
Very happy to see our paper on Equivariant Flow-based sampling for lattice gauge theory out in Physical Review Letters! journals.aps.org/prl/abstract/1… cc: Tej Kanwar, Denis Boyda, @DaniloJRezende @KyleCranmer @sracaniere, Dan Hackett, Phiala Shanahan