Lightnews — Scholar-powered news

Reposted by Antoine Moulin

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 6d

We're finally out of stealth: percepta.ai
We're a research / engineering team working together in industries like health and logistics to ship ML tools that drastically improve productivity. If you're interested in ML and RL work that matters, come join us 😀

Percepta | A General Catalyst Transformation Company

Transforming critical institutions using applied AI. Let's harness the frontier.

percepta.ai

7 18 120

Reposted by Antoine Moulin

Sonia Mazelet @soniamazelet.bsky.social · 9d

I am happy to share that our paper "Unsupervised Learning for Optimal Transport plan prediction between unbalanced graphs" was accepted at Neurips 2025 ! 🥳

Huge thanks to my co-authors @rflamary.bsky.social and Bertrand Thirion !

arxiv.org/abs/2506.12025

(1/5)

Unsupervised Learning for Optimal Transport plan prediction between unbalanced graphs

Optimal transport between graphs, based on Gromov-Wasserstein and other extensions, is a powerful tool for comparing and aligning graph structures. However, solving the associated non-convex optimizat...

arxiv.org

1 4 29

Reposted by Antoine Moulin

ARLET @arletworkshop.bsky.social · Aug 28

News 🎉 We’re thrilled to announce our final panelist: David Silver!
Don’t miss David and our amazing lineup of speakers—submit your latest RL work to our NeurIPS workshop.
📅 Extended deadline: Sept 2 (AoE)

2 3

Reposted by Antoine Moulin

ARLET @arletworkshop.bsky.social · Aug 18

We've extended the deadline for our workshop's calls for papers/ideas! Submit your work by August 29 AoE. Instructions on the website: arlet-workshop.github.io/neurips2025/...

Call for Papers | ARLET

A simple, whitespace theme for academics. Based on [*folio](https://github.com/bogoli/-folio) design.

arlet-workshop.github.io

2 3

Reposted by Antoine Moulin

ARLET @arletworkshop.bsky.social · Aug 5

The OpenReview link for our calls (for papers and ideas) is available, submit here: openreview.net/group?id=Neu...

We look forward to receiving your submissions!

1 2

Antoine Moulin @antoine-mln.bsky.social · Jul 28

last year's edition was so much fun I'm really looking forward to this one!! join us in San Diego :))

ARLET @arletworkshop.bsky.social · Jul 28

Delighted to announce that the second edition of our workshop has been accepted as a #NeurIPS2025 workshop!

4

Antoine Moulin @antoine-mln.bsky.social · Jun 19

Was it recorded? 🤔

1 1

Reposted by Antoine Moulin

RL Theory Virtual Seminars @rl-theory.bsky.social · Jun 9

Join us for Nneka's presentation tomorrow! Last talk before the summer break.

3 9

Reposted by Antoine Moulin

RL Theory Virtual Seminars @rl-theory.bsky.social · Jun 2

Join us tomorrow for Dave's talk! He will present his recent work on randomised exploration, which received an outstanding paper award at ALT 2025 earlier this year.

1 4

Antoine Moulin @antoine-mln.bsky.social · May 27

link: arxiv.org/abs/2505.19946

Inverse Q-Learning Done Right: Offline Imitation Learning in $Q^π$-Realizable MDPs

We study the problem of offline imitation learning in Markov decision processes (MDPs), where the goal is to learn a well-performing policy given a dataset of state-action pairs generated by an expert...

arxiv.org

3

Antoine Moulin @antoine-mln.bsky.social · May 27

new preprint with the amazing @lviano.bsky.social and @neu-rips.bsky.social on offline imitation learning! learned a lot :)

when the expert is hard to represent but the environment is simple, estimating a Q-value rather than the expert directly may be beneficial. lots of open questions left though!

1 3 18

Reposted by Antoine Moulin

Dimitri Meunier @dimitrimeunier.bsky.social · May 26

🚨 New paper accepted at SIMODS! 🚨
“Nonlinear Meta-learning Can Guarantee Faster Rates”

arxiv.org/abs/2307.10870

When does meta learning work? Spoiler: generalise to new tasks by overfitting on your training tasks!

Here is why:
🧵👇

Nonlinear Meta-Learning Can Guarantee Faster Rates

Many recent theoretical works on \emph{meta-learning} aim to achieve guarantees in leveraging similar representational structures from related tasks towards simplifying a target task. The main aim of ...

arxiv.org

2 7 9

Reposted by Antoine Moulin

Dylan Foster 🐢 @djfoster.bsky.social · May 26

Dhruv Rohatgi will be giving a lecture on our recent work on comp-stat tradeoffs in next-token prediction at the RL Theory virtual seminar series (rl-theory.bsky.social) tomorrow at 2pm EST! Should be a fun talk---come check it out!!

1 5 11

Reposted by Antoine Moulin

Gergely Neu @neu-rips.bsky.social · May 26

new work on computing distances between stochastic processes ***based on sample paths only***! we can now:
- learn distances between Markov chains
- extract "encoder-decoder" pairs for representation learning
- with sample- and computational-complexity guarantees
read on for some quick details..
1/n

1 10 37

Reposted by Antoine Moulin

Ferenc Huszár @inference.vc · May 22

A new blog post with intuitions behind continuous-time Markov chains, a building block of diffusion language models, like @inceptionlabs.bsky.social's Mercury and Gemini Diffusion. This post touches on different ways of looking at Markov chains, connections to point processes, and more.

Discrete Diffusion: Continuous-Time Markov Chains

A tutorial explaining some key intuitions behind continuous time Markov chains for machine learners interested in discrete diffusion models: alternative representations, connections to point processes...

www.inference.vc

1 5 22

Antoine Moulin @antoine-mln.bsky.social · May 22

oh the inference blog posts are back 🥰

1

Reposted by Antoine Moulin

arXiv cs.LG Machine Learning @cslg-bot.bsky.social · May 21

Mattes Mollenhauer, Nicole M\"ucke, Dimitri Meunier, Arthur Gretton: Regularized least squares learning with heavy-tailed noise is minimax optimal https://arxiv.org/abs/2505.14214 https://arxiv.org/pdf/2505.14214 https://arxiv.org/html/2505.14214

1 6 6

Reposted by Antoine Moulin

RL Theory Virtual Seminars @rl-theory.bsky.social · May 20

Later today, Sikata and Marcel will talk about their recent work on oracle-efficient RL with ensembles. Join us!

4 6

Reposted by Antoine Moulin

Edouard Leurent @eleurent.bsky.social · May 20

Excited to share what I've been up to: bringing text diffusion to Gemini!

Diffusion models are _fast_, and hold immense promise to challenge autoregressive models as the de facto standard for language modeling.

1 2 15

Antoine Moulin @antoine-mln.bsky.social · May 16

omg thanks

Reposted by Antoine Moulin

Clément Canonne @ccanonne.github.io · May 10

The tutorials, workshops, and community events for #COLT2025 have been announced!

Exciting topics, and impressive slate of speakers and events, on June 30! The workshops have calls for contributions (⏰ May 16, 19, and 25): check them out!
learningtheory.org/colt2025/ind...

Community events and tutorials, list from the website

2 7 20

Reposted by Antoine Moulin

Dylan Foster 🐢 @djfoster.bsky.social · May 9

Announcing the first workshop on Foundations of Post-Training (FoPT) at COLT 2025!

📝 Soliciting abstracts/posters exploring theoretical & practical aspects of post-training and RL with language models!

🗓️ Deadline: May 19, 2025

1 6 17

Reposted by Antoine Moulin

Gergely Neu @neu-rips.bsky.social · May 6

looking forward to giving a talk at the 2025 GHOST day in the beautiful city of Poznan (Poland) --- hope to see you there this weekend!

1 1 7

Reposted by Antoine Moulin

Dylan Foster 🐢 @djfoster.bsky.social · May 3

Is Best-of-N really the best we can do for language model inference?

New paper (appearing at ICML) led by the amazing Audrey Huang (ahahaudrey.bsky.social) with Adam Block, Qinghua Liu, Nan Jiang, and Akshay Krishnamurthy (akshaykr.bsky.social).

1/11

1 5 22

Reposted by Antoine Moulin

Arthur Gretton @arthurgretton.bsky.social · May 2

Spectral Representation for Causal Estimation with Hidden Confounders
at #AISTATS2025

A spectral method for causal effect estimation with hidden confounders, for instrumental variable and proxy causal learning
arxiv.org/abs/2407.10448

Haotian Sun, @antoine-mln.bsky.social, Tongzheng Ren, Bo Dai

1 3 3