rl-agent

[Up-to-date] A curated list of resources on graph-empowered agents and agent-facilitated graph learning (Graphs Meet Agents & Agentic Graph Engineering).

Non-deterministic, human-like mouse and keyboard automation powered by reinforcement learning.

Deep Reinforcement Learning Texas Hold'em AI / 德州AI/深度强化学习德州扑克AI / 深度強化學習德州撲克AI - CFR, DQN, AlphaZero-style training

A smart traffic light management dashboard using SUMO and reinforcement learning to simulate traffic flow, optimize signal timing, and improve vehicle movement at intersections.

Train SLM to use Tools with RL

Reinforcement learning Tetris bot

A robust Reinforcement Learning environment for the oink games. Compatible with OpenAI Gym interface.

An interactive browser-based playground to learn reinforcement learning — from bandits to policy gradients. Watch 13 algorithms learn in real time across 4 environments (Bandit, GridWorld, CartPole, Rocket Landing), tune hyperparameters, and follow a built-in 10-chapter course. No backend, no account — just open and learn.

Sprite Garden

Deep Q-Network (DQN) agent trained to play Flappy Bird using PyTorch & Gymnasium — with experience replay, target network, and epsilon-greedy exploration.

A robust reinforcement learning framework for voicebot turn-level decision optimization. Utilizes offline RL techniques to learn from historical, noisy, and delayed feedback signals without deploying exploratory policies, ensuring safe and stable conversational improvement.

A Pong game which you can actually play with a sophisticated RL Agent !!

Bitcoin trading agent using Deep Q-Learning and synthetic market scenarios.

A Blackjack environment and reinforcement learning agent that learns optimal play from scratch. It includes a full game simulator, Monte Carlo basic strategy solver, and tabular Q-learning.

A reinforcement learning agent that learns to play Pacman in a custom environment with a visual GUI.

Automated Text Generation Using English Grammar (AI but AI is a Lie 😂, its Mathematics. AI, LLMs, because these keywords are trending 😂)

🏓 A self-learning Pong agent blending value-based and policy-based RL — playable in your terminal.

A high-density browser-based evolutionary ecosystem simulation combining swarm behaviour, reinforcement-learning-style commander control, environmental pressure, emergent cooperation, resource competition, pathogen dynamics, and research-inspired quality-diversity tracking.

Reinforcement learning framework for portfolio optimization using PPO agents to learn dynamic ETF allocation strategies