Foundations · MLD-1012
Reinforcement learning
Learning action-selection behavior from interaction and reward signals.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Train a game-playing policy using game outcomes.
Original caveat
Reward is a designed signal, not necessarily the intended goal.
Context and aliases
Level: Foundational. Stages: Theory · Training · Inference.
Tags: ML · DL · SLM · LLM.
Aliases: RL.
Related concepts
Technology connections
- Brax — JAX-based reinforcement-learning training utilities with legacy physics/environment components.
- CleanRL — Single-file implementations of deep RL algorithms.
- Gymnasium — Standard API and environment ecosystem for reinforcement learning.
- MuJoCo — Physics engine for articulated-body simulation.
- NVIDIA Isaac Sim — Robotics simulation platform with rendering and sensor tools.
- PettingZoo — API and environment collection for multi-agent reinforcement learning.
- RLlib — Distributed reinforcement-learning library in Ray.
- Stable-Baselines3 — Tested implementations of common RL algorithms in PyTorch.
- TorchRL — PyTorch components for reinforcement learning.
Original references
- Deep Learning: online textbook ↗ · Consulted
- Google Machine Learning Glossary ↗ · Consulted
- Original reference URL ↗