AI agent taught itself to mine crypto during training from boingboing rss feed

AI agent taught itself to mine crypto during training
boingboing.net/2026/03/23/ai-agent-taug...
#AI_safety #artificial_intelligence #cryptocurrency #reinforcement_learning #Technology #boingboing

DeepSeek R1 proved you can teach a model to reason through pure reinforcement learning, no supervised fine-tuning. 671B params, 37B active. Matches o1 performance at 15-50% cost. MIT licensed. Turns out scale isn't everything.

AgiBot Celebrates Groundbreaking Deployment of Reinforcement Learning in Robotics AgiBot has achieved a significant milestone with its Real-World Reinforcement Learning system, marking a new era in industrial robotics.

AgiBot Celebrates Groundbreaking Deployment of Reinforcement Learning in Robotics #Shanghai #USA #Robotics #AgiBot #Reinforcement_Learning

AgiBot Achieves Landmark Success in Real-World Robotics with Reinforcement Learning AgiBot has marked a critical milestone by successfully deploying its Real-World Reinforcement Learning system in industrial robotics, showcasing a leap in intelligent automation.

AgiBot Achieves Landmark Success in Real-World Robotics with Reinforcement Learning #Robotics #AgiBot #Reinforcement_Learning

Celebrating the 2024 ACM A.M. Turing Award Winners: Pioneers in AI Technology Development Discover how Andrew G. Barto and Richard S. Sutton transformed the field of AI through their groundbreaking work in reinforcement learning, earning the prestigious 2024 ACM A.M. Turing Award.

Celebrating the 2024 ACM A.M. Turing Award Winners: Pioneers in AI Technology Development #USA #New_York #Reinforcement_Learning #ACM_Award #Barto_Sutton

Xiao-I's Strategic Moves in AI: DeepSeek Insights and U.S. Expansion Plans Xiao-I Corporation discusses its advancements in AI technology with the launch of its cost-efficient Hua Zang LLM and plans for expansion in the U.S.

Xiao-I's Strategic Moves in AI: DeepSeek Insights and U.S. Expansion Plans #United_States #Piscataway #Xiao-I #Hua_Zang_LLM #Reinforcement_Learning

Diminishing Return of Value Expansion Methods Model-based reinforcement learning aims to increase sample efficiency, but the accuracy of dynamics models and the resulting compounding errors are often seen as key limitations. This paper empiricall...

Diminishing Return of Value Expansion Methods

Authors: Daniel Palenicek, Michael Lutter, João Carvalho, Daniel Dennert, Faran Ahmad, and Jan Peters Fellow

pre-print -> arxiv.org/abs/2412.20537

#rl #reinforcement_learning #modelbased_rl #value_expension

Overview from Robot Learning with Super-Linear Scaling

Results from Robot Learning with Super-Linear Scaling

Robot Learning with Super-Linear Scaling

Authors: M. Torne, A. Jain, J. Yuan, V. Macha, L. Ankile, A. Simeonov, P. Agrawal, A. Gupta

pre-print -> arxiv.org/abs/2412.017...
website -> casher-robot-learning.github.io/CASHER/

#robotics #rl #reinforcement_learning #data_generation #real2sim2real

Enhancing Exploration with Diffusion Policies in Hybrid Off-Policy RL: Application to Non-Prehensile Manipulation

Authors: Huy Le et al.

pre-print -> arxiv.org/abs/2411.14913
website -> leh2rng.github.io/hydo

#robotics #rl #reinforcement_learning #nonprehensile #manipulation #diffusion #entropy

I am amazed by the amount of #Robotics and #RL #reinforcement_learning people here!

Opinion | Press Pause on the Silicon Valley Hype Machine A.I. is looking less like an all-powerful being and more like an unreliable intern.

#AIhype gets called out in the The New York Times Sunday opinion section: "it’s looking less like an all-powerful being and more like a bad intern" Fortunately my #datascience work is using old school stuff like #statistics and #reinforcement_learning not chatbots! www.nytimes.com/2024/05/15/o...