Marco Max Fiandri
@marcomaxfiandri.bsky.social
Bandits, Probability and gags
Reposted by Marco Max Fiandri
Had an amazing time presenting my research @cohereforai.bsky.social yesterday 🎤
In case you could not attend, feel free to check it out 👉
youtu.be/RCA22JWiiY8?...
In case you could not attend, feel free to check it out 👉
youtu.be/RCA22JWiiY8?...
Théo Vincent - Optimizing the Learning Trajectory of Reinforcement Learning Agents
YouTube video by Cohere
youtu.be
July 19, 2025 at 7:42 AM
Had an amazing time presenting my research @cohereforai.bsky.social yesterday 🎤
In case you could not attend, feel free to check it out 👉
youtu.be/RCA22JWiiY8?...
In case you could not attend, feel free to check it out 👉
youtu.be/RCA22JWiiY8?...
Here we go again with shameless selfpromotion, now revised to include literally any Restless setting providing state-of-the-art guarantees for the quite challenging Beta-Binomial case.
Marco Fiandri, Alberto Maria Metelli, Francesco Trov\`o
Sliding-Window Thompson Sampling for Non-Stationary Settings
https://arxiv.org/abs/2409.05181
Sliding-Window Thompson Sampling for Non-Stationary Settings
https://arxiv.org/abs/2409.05181
June 17, 2025 at 12:38 PM
Here we go again with shameless selfpromotion, now revised to include literally any Restless setting providing state-of-the-art guarantees for the quite challenging Beta-Binomial case.
(Shameless self promotion)-There are few interesting technical insights for those interested in TS
Marco Fiandri, Alberto Maria Metelli, Francesco Trov\`o
Thompson Sampling-like Algorithms for Stochastic Rising Bandits
https://arxiv.org/abs/2505.12092
Thompson Sampling-like Algorithms for Stochastic Rising Bandits
https://arxiv.org/abs/2505.12092
May 22, 2025 at 4:28 PM
(Shameless self promotion)-There are few interesting technical insights for those interested in TS
Reposted by Marco Max Fiandri
fool me once, shame on you
assume fool me n times, shame on you
suppose you fool me n+1 times,
assume fool me n times, shame on you
suppose you fool me n+1 times,
April 3, 2025 at 9:06 AM
fool me once, shame on you
assume fool me n times, shame on you
suppose you fool me n+1 times,
assume fool me n times, shame on you
suppose you fool me n+1 times,
Reposted by Marco Max Fiandri
Very happy to announce that iterated Q-Network (i-QN) has been published in TMLR 🎉
i-QN learns several Bellman iterations in parallel instead of learning them sequentially via repeated target updates ✨ This directly translates to performance improvements on the Atari and MuJoCo benchmarks 🚀
i-QN learns several Bellman iterations in parallel instead of learning them sequentially via repeated target updates ✨ This directly translates to performance improvements on the Atari and MuJoCo benchmarks 🚀
Iterated $Q$-Network: Beyond One-Step Bellman Updates in Deep Reinforcement Learning
Théo Vincent, Daniel Palenicek, Boris Belousov, Jan Peters, Carlo D'Eramo
Action editor: Pablo Castro
https://openreview.net/forum?id=Lt2H8Bd8jF
#reinforcement #iterative #iterations
Théo Vincent, Daniel Palenicek, Boris Belousov, Jan Peters, Carlo D'Eramo
Action editor: Pablo Castro
https://openreview.net/forum?id=Lt2H8Bd8jF
#reinforcement #iterative #iterations
March 24, 2025 at 2:49 PM
Very happy to announce that iterated Q-Network (i-QN) has been published in TMLR 🎉
i-QN learns several Bellman iterations in parallel instead of learning them sequentially via repeated target updates ✨ This directly translates to performance improvements on the Atari and MuJoCo benchmarks 🚀
i-QN learns several Bellman iterations in parallel instead of learning them sequentially via repeated target updates ✨ This directly translates to performance improvements on the Atari and MuJoCo benchmarks 🚀
Reposted by Marco Max Fiandri
Iterated $Q$-Network: Beyond One-Step Bellman Updates in Deep Reinforcement Learning
Théo Vincent, Daniel Palenicek, Boris Belousov, Jan Peters, Carlo D'Eramo
Action editor: Pablo Castro
https://openreview.net/forum?id=Lt2H8Bd8jF
#reinforcement #iterative #iterations
Théo Vincent, Daniel Palenicek, Boris Belousov, Jan Peters, Carlo D'Eramo
Action editor: Pablo Castro
https://openreview.net/forum?id=Lt2H8Bd8jF
#reinforcement #iterative #iterations
February 23, 2025 at 3:07 PM
Iterated $Q$-Network: Beyond One-Step Bellman Updates in Deep Reinforcement Learning
Théo Vincent, Daniel Palenicek, Boris Belousov, Jan Peters, Carlo D'Eramo
Action editor: Pablo Castro
https://openreview.net/forum?id=Lt2H8Bd8jF
#reinforcement #iterative #iterations
Théo Vincent, Daniel Palenicek, Boris Belousov, Jan Peters, Carlo D'Eramo
Action editor: Pablo Castro
https://openreview.net/forum?id=Lt2H8Bd8jF
#reinforcement #iterative #iterations
Reposted by Marco Max Fiandri
One year anniversary of our local @twiml.bsky.social Tübingen Women in ML and our community is growing bigger and stronger. 100 registrations last week! Local workshops really have impact. Looking forward to our next event in autumn.
Friday was amazing! A huge thank you to all the participants and speakers who made this event possible. It’s amazing to see this community growing! A special thanks to the @ml4science.bsky.social for their support. Looking forward to seeing you at the next event!
March 17, 2025 at 8:12 PM
One year anniversary of our local @twiml.bsky.social Tübingen Women in ML and our community is growing bigger and stronger. 100 registrations last week! Local workshops really have impact. Looking forward to our next event in autumn.
Me irl when I mix my daily dose of caffeine pre-overleaf with ketoprofen
March 11, 2025 at 2:03 PM
Me irl when I mix my daily dose of caffeine pre-overleaf with ketoprofen
Reposted by Marco Max Fiandri
We’re celebrating #InternationalWomensDay early by shining a spotlight on some of the incredible women who have shaped the world of mathematics! 💡✨ Learn more about their contributions and impact:
11 Famous Women Mathematicians and Their Incredible Contributions! — Mashup Math
Celebrate Women's History by learning about these famous women mathematicians and scientists and their amazing contributions. A full biography of each famous female mathematician is included!
www.mashupmath.com
March 7, 2025 at 7:52 PM
We’re celebrating #InternationalWomensDay early by shining a spotlight on some of the incredible women who have shaped the world of mathematics! 💡✨ Learn more about their contributions and impact:
Reposted by Marco Max Fiandri
Meet the recipients of the 2024 ACM A.M. Turing Award, Andrew G. Barto and Richard S. Sutton! They are recognized for developing the conceptual and algorithmic foundations of reinforcement learning. Please join us in congratulating the two recipients! bit.ly/4hpdsbD
March 5, 2025 at 2:26 PM
Meet the recipients of the 2024 ACM A.M. Turing Award, Andrew G. Barto and Richard S. Sutton! They are recognized for developing the conceptual and algorithmic foundations of reinforcement learning. Please join us in congratulating the two recipients! bit.ly/4hpdsbD