It's been an uneventful end of week, so here's something interesting to fill the slow news cycle! Our next TCS+ talk will be by Kasper Green Larsen, on Nov 29 — on how Bagging is an Optimal PAC Learner. Mark your calendar... more details coming soon!

Interested in helping to organize NeurIPS Conference 2024? Let us know here: docs.google.com/forms/u/1/d/e/…

Our work "Data-Driven Online Model Selection With Regret Guarantees " will be @ #AISTATS2024! Our algorithms satisfy data dependent model selection bounds and are very simple and beautiful! Joint work with Christoph Dann Chris Dann and Claudio Gentile. arxiv.org/abs/2306.02869

Hi everybody - excited to announce the "ITALT" symposium between ITA and ALT that I'm organizing in San Diego together with daniel hsu, Claire Vernade and Alon Orlitsky to bridge the information theory and learning theory communities!

The Complexity of Sequential Prediction in Dynamical Systems ift.tt/hyp4mbx

Apple Tasting: Combinatorial Dimensions and Minimax Rates. arxiv.org/abs/2310.19064

Bandit-Feedback Online Multiclass Classification: Variants and Tradeoffs. arxiv.org/abs/2402.07453

On Limitations of the Transformer Architecture ift.tt/bR6gyfD

The NeurIPS 2024 Call for Papers is now live! Abstracts due May 15, 2024 01:00 PM PDT Full paper submission deadline is May 22, 2024 01:00 PM PDT buff.ly/3TDiQxI

Optimal Multi-Distribution Learning ift.tt/I3W68xd

The Real Price of Bandit Information in Multiclass Classification ift.tt/FGvI2OE

Another new blog, this time on the optimal sample complexity of PAC learning, by Ishaq Aden-Ali Yeshwanth Cherapanamjeri Abhishek Shetty and Nikita Zhivotovskiy. Check it out: let-all.com/blog/2024/06/0…

Last reminder that the "early bird" ticket sales for #COLT2024 ends tomorrow, on June 14! Shipra Agrawal and Aaron Roth put together an awesome program. See learningtheory.org/colt2024/sched… Please repost for visibility. Hope to see you in Edmonton!

A Theory of Interpretable Approximations ift.tt/0URerAk

Is Efficient PAC Learning Possible with an Oracle That Responds 'Yes' or 'No'? by Noah Golowich and Constantinos Daskalakis Beautiful results and techniques! arxiv.org/abs/2406.11667

Credit Attribution and Stable Compression ift.tt/pAag8F5

Ramsey Theorems for Trees and a General 'Private Learning Implies Online Learning' Theorem ift.tt/f9ExZXq

Error Bounds for Learning Fourier Linear Operators ift.tt/XK1NEGg

Generation through the lens of learning theory ift.tt/phnf1XY

Sample Compression Scheme Reductions ift.tt/zHNIX8L