"pytorch-optimizer -- collections of ready to use optimization algorithms for PyTorch" github.com/jettify/pytorc…

This ultra-small generator is highly portable and works even with shallow, slow moving water. Already tested with powering street lights, it promises to allow people in the world’s remote regions to generate their own electricity buff.ly/2U6yeEc

dspace.mit.edu/bitstream/hand… How to do Research At the MIT AI Lab (1988). Almost all advices are still valid more than three decades later. Highly recommended.

✈️ For our flight-control nerds 🗺 The Map of Control Theory by Brian Douglas

excited to announce a new series of virtual seminars on ~~~REINFORCEMENT LEARNING THEORY~~~ we've set this up with Ciara Pike-Burke and Csaba Szepesvari to keep track of all the advances of this fast-paced field. hope others will also find it useful! sites.google.com/view/rltheorys…

Battling COVID-19 | Think outside the box | Think beyond contact tracing. A method like Safe Blues can help with quick detection of the second wave. Safe Blues paper: safeblues.org

Excited to introduce the AI Economist: Extends ideas from Reinforcement Learning for tackling inequality through learned tax policy design. The framework optimizes productivity and equality. Blog: blog.einstein.ai/the-ai-economi… Paper: arxiv.org/abs/2004.13332 Q&A: salesforce.com/company/news-p…

We just released our new paper: A Finite Time Analysis of Two Time-Scale Actor Critic Methods (arxiv.org/abs/2005.01350). We for the first time show that two time-scale actor-critic methods can find an eps-stationary point within eps^{-2.5} sample complexity.

Imagine seeing a 20 storey tower flying into space and then landing itself upright. This will never stop blowing my mind. 🤯 Elon Musk SpaceX

Can you beat this robot? 🤖🎾 #AI #MachineLearning #Sports #Robotics Mike Quindazzi Harold Sinnott 📱 Rajat Shrimal Sen. Sally Eaves Spiros Margaris Jean-Baptiste Lefevre Helen Yu Kohei Kurihara - Privacy for all together 🌍 Dr. Marcell Vollmer #StaySafe #CES2026 Kohei Kurihara - Privacy for all together 🌍 Christine Boursin 🌐 Ronald van Loon Andreas Staub Helene Li ✨🍀💜 x.com/Paula_Piccard/…

Researchers and Founders blog.samaltman.com/researchers-an…

"Performing a good experiment is as demanding as proving a new theorem" (Hans-Paul Schwefel)

What's new, Atlas? youtu.be/fRj34o4hN4I via YouTube

Average-reward model-free reinforcement learning: a systematic review and literature mapping deepai.org/publication/av… by Vektor Dewanto et al. including Marcus Gallagher #ReinforcementLearning #ArtificialIntelligence

A nearly Blackwell-optimal policy gradient method deepai.org/publication/a-… by Vektor Dewanto and Marcus Gallagher #ReinforcementLearning #ComputerScience

Examining average and discounted reward optimality criteria in reinforcement learning deepai.org/publication/ex… by Vektor Dewanto and Marcus Gallagher #ReinforcementLearning #ComputerScience

The Statistics with The Julia Language book is published! statisticswithjulia.org

Approximate discounting-free policy evaluation from transient and recurrent states deepai.org/publication/ap… by Vektor Dewanto and Marcus Gallagher #Statistics #Estimator

Will Kurt: "Close up rigor can be very confusing, but with perspective rigor adds clarity." So true... thanks Will Kurt

Like it: the opt path, from twosigma.com/wp-content/upl…