Featured topics
Topic Reinforcement Learning
2 articles
AI DevelopmentWe Taught Robots "Sumo"—and They Mastered Lariats, Cossack Dancing, and Aerial Combat
We trained two robots to fight sumo in NVIDIA Isaac Sim and tracked their evolution from Epoch 1 to 100. The experiment explores reinforcement learning, Sim-to-Real, domain randomization, and the unexpected strategies the AI discovered—from lariats and kicks to low stances and aerial combat.
Toshihiko Nagaoka
AI DevelopmentAI Economist: Reinforcement Learning in Economics
Experimenting with the AI Economist API to simulate free markets and communism using Reinforcement Learning (RL) agents.
Vitalify Asia Team
