Michele Caprio

And for @EdgarDobriban of course!

Very well deserved! Fun fact: I would never have been a postdoc at @Penn if it weren’t for Yuting :)

Very excited to share that our ICML workshop proposal got accepted! See you in Seoul :)

Very fun work led by the amazing Anita, @Chau9991, and @masaki_adachi on how LLMs verbalize higher-order uncertainty, using tools from imprecise probabilities. Let us know what you think! arxiv.org/abs/2603.10396

A very fun project, joint with the star Nina Gottschling (ornl.gov/staff-profile/…) @ORNL :) Let us know what you think!

We proved Hoeffding-style concentration bounds for exchangeable random variables, extending previous work in the exchangeable setting that was restricted to population means! arxiv.org/abs/2603.10190

Marius Furter, Yujun Huang, Gioele Zardini: Composable Uncertainty in Symmetric Monoidal Categories for Design Problems arxiv.org/abs/2603.09430 arxiv.org/pdf/2603.09430 arxiv.org/html/2603.09430

@ISBA_events, @j_ISBA

Very nice blog post by one of our CDT PhD Students Abhishek (github.com/abjodas) on Bayesian Optimization, give it a read! abjodas.quarto.pub/aicdt-blog/bo-…

A meditation on the meaning of mathematical beauty in the age of AI, by @rita_ahmadi aeon.co/essays/how-sho…

Bringing this back

@_onionesque Looks like a great collection!

Barely started unboxing books and already hurt my back. This is gonna take months.

Category theory is a mathematical way to formalize the relationships between objects. It could provide a powerful method for modeling complex, interconnected systems. Read @nattyover's new Qualia column: quantamagazine.org/can-the-most-a…

Comments and thoughts are very welcome (especially from the @SIPTAtweets community)! 4/n, n=4

In particular, we leverage Bakry-Émery curvature to turn geometric information about the state space into ergodic bounds under model uncertainty. We also discuss some connections with AI and computer vision. 3/n

Excited to share our new paper! arxiv.org/abs/2405.00081 We study how to represent uncertainty in continuous-time Markov processes through collections of Markov semigroups (imprecise Markov semigroups), ... 1/n

@Almost_Sure Liverpool? :)

Left the house this morning, and the sky had turned a weird blue colour. Stay safe out there!

One reason why I'm a theorist @Dirk_Englund #SIPQNP @RLEatMIT

🚨 New Textbook on Conformal Prediction 🚨 arxiv.org/abs/2411.11824 “The goal of this book is to teach the reader about the fundamental technical arguments that arise when researching conformal prediction and related questions in distribution-free inference. Many of these proof strategies, especially the more recent ones, are scattered among research papers, making it difficult for researchers to understand where to look, which results are important, and how exactly the proofs work. We hope to bridge this gap by curating what we believe to be some of the most important results in the literature and presenting their proofs in a unified language, with illustrations, and with an eye towards pedagogy.” We are looking for feedback — and this is only a draft, with Part 4 coming soon! Please reach out! With Rina Foygel Barber and @stats_stephen