home.social

#deepreinforcementlearning — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #deepreinforcementlearning, aggregated by home.social.

  1. Autonomy Talks - Georgia Chalvatzaki: Shaping #Robotic Assistance through Structured #Robot #Learning: youtube.com/watch?v=e0aQC3C8P7 #robotics #machinelearning

    Around 12:30 they present the training of a model-free #MDP #deepreinforcementlearning using a model-based #ai #planner #aiplanner. Indeed it drastically boosts the training.

    The general idea is to guide an implicit model using a model-based approximation, and it works also for assembly tasks, computer vision, pick and place…

  2. AI helps 3D printers “write” with coiling fluid ropes like Jackson Pollock - Enlarge / Jackson Pollock working in his Long Island studio adjacent to... - arstechnica.com/?p=1980704 #deepreinforcementlearning #machinelearning #jacksonpollock #fluiddynamics #scienceandart #3dprinting #science #physics

  3. Faster sorting #algorithms discovered using #DeepReinforcementLearning | #Nature

    "Here we show how #ArtificialIntelligence can go beyond the current state of the art by discovering hitherto unknown routines. To realize this, we formulated the task of finding a better sorting routine as a single-player game. We then trained a new deep #ReinforcementLearning agent, #AlphaDev, to play this game. AlphaDev discovered small sorting algorithms from scratch that outperformed previously known human benchmarks. These algorithms have been integrated into the #LLVM standard C++ sort library3. This change to this part of the sort library represents the replacement of a component with an algorithm that has been automatically discovered using reinforcement learning."

    nature.com/articles/s41586-023