pyboy

Playing Pokemon Red using TypeSafe Jev

Autonomous Pokémon Red agent using local Qwen3-VL and PyBoy. Zero API calls, zero cost, pure local VLM reasoning and computer vision.

Gameboy (Color) Environments in Gymnasium.

ZeDMD-compatible network DMD for Raspberry Pi — drives S-PWM HUB75 panels (FM6373 and similar) that ZeDMD cannot, and between games it stays a living-room display: clock, flight radar, ISS pass alerts, now playing, calendar, Doom and Game Boy, with two-way Home Assistant integration over MQTT

Local Pokemon Red LLM-agent harness with live gameplay UI, structured traces, save states, replay, and turn-by-turn agent observability.

A real fruit fly connectome (MaleCNS v1.0, 166,700 neurons) running inside Game Boy games via PyBoy

Work in progress: agents playing the real Pokémon Blue ROM on PyBoy, with every badge, catch and evolution verified in cartridge RAM.

🎮🌟A Reinforcement Learning (RL) project using Double Deep Q-Network (DDQN) to train an agent to play Kirby's Dreamland.

Trains a deep Q-network to play Pokémon Blue from the screen, through PyBoy and PyTorch.

Twitch-plays-Pokemon style bot for Slack using PyBoy

Completed reproducible Pokémon Red study: 8.24M self-generated actions reached Route 1, but frozen evaluation found zero durable skills.

Agentic PokeSim fork: FastAPI + PyBoy backend with an OpenAI-compatible vision agent and per-game AI selection

ROM-free OpenRappter agent for a local Copilot-powered Pokemon Red playthrough with PyBoy

A learning experiment exploring how to teach an AI model how to play the game Tetris for GameBoy using StableBaselines3 and PyBoy.

AI that learns to play Pokemon Red: an LLM discovers the controls and designs rewards, a small local CNN trained with PPO plays on parallel headless Game Boy emulators (PyBoy).

Hold-aware DQN training for Pandora's Blocks through PyBoy

A fixed agent harness for Pokémon Red where the model is the only variable: same tools, memory, prompt and observation format across 5 models from 4 vendors, turn-matched and eval-integrity gated. Live leaderboard with per-seed spreads and every exclusion.

I wanted to make an LLM play Pokemon, comparable to what Microsoft and Anthropic had their AIs do. I decided to switch to Deadeus for now, as an open source alternative.

独立复现 Pokemon Red RL (PPO / neroRL poke_red) + 1,000,000 步监督训练实验与固定协议独立评估。不含 ROM/存档/权重/录像。

AI Gym environment for Castlevania: The Adventure (Game Boy), using the PyBoy emulator.