RL Demo Suite

Four reinforcement-learning agents, each optimizing an immutable LLM-consensus reward on GenLayer. Pick an agent to watch it act, get judged, and learn.