yeshenpy/Awesome-Evolutionary-Reinforcement-Learning

Research Papers and Code Repository on the Integration of Evolutionary Algorithms and Reinforcement Learning

What it solves

Researchers entering Evolutionary Reinforcement Learning (ERL) face a scattered field: many hybrid algorithms, a taxonomy of approaches, and often unclear which papers have usable code. This repository solves that by serving as the official companion to a comprehensive survey, condensing the field into an organized index so readers can find, understand, and reproduce representative methods without hunting across dozens of sources.

How it works

The repository organizes ERL into three major directions (EA assists RL, RL assists EA, and EA and RL collaborate) plus sixteen research branches, then extends that taxonomy to twenty-one maintained branches. It indexes 270 canonical algorithm works and tracks 80 with verified public code. A second layer provides 86 independently reviewed bilingual (Chinese/English) reading-guide mappings covering 54 works, with Core explanations (substantial, work-specific) and Extended ones (concise overviews). The code-first reading list highlights 31 papers that have verified public implementations, labeled as author-maintained or community reimplementations. The repository also tracks recent and emerging directions and maintains a changelog tied to survey version updates.

Who it’s for

This is for researchers, students, and practitioners who want to study, reproduce, or extend hybrid evolutionary-reinforcement-learning methods. It is especially useful for those who need a clear taxonomy, verified code links, and bilingual explanations to quickly move from the survey paper to implementation and new work.

Highlights

  • Official companion to the survey "Bridging Evolutionary Algorithms and Reinforcement Learning" (IEEE TEVC 2025, arXiv v5).
  • Indexes 270 canonical algorithm works across three directions and twenty-one branches.
  • 80 canonical works have verified public code, with 31 in a code-first reading list.
  • Includes 86 independently reviewed bilingual (中文/English) reading-guide mappings covering 54 works, with Core and Extended levels.
  • Provides a taxonomy figure, integration-pattern figure, and clear guidance on survey versions (IEEE for citation, arXiv v5 for latest content, repository for implementation tracking).
  • Actively maintained with a changelog and contributions welcome.

Related

  • Project
    ML4ITS/mtad-gat-pytorch

    A PyTorch implementation of MTAD-GAT for multivariate time-series anomaly detection using Graph Attention Networks and GRU-based reconstruction.

  • Project
    next-state/open-dreamer

    Open Dreamer is a JAX/Flax implementation of the Dreamer 4 world‑model pipeline, providing training code for a video tokenizer and an action‑conditioned latent dynamics model on Minecraft gameplay data, plus tools for rollout generation and FVD evaluation. It includes a live in‑browser demo and a separate inference repo for running trained checkpoints.

  • Project
    SciML/SciMLSensitivity.jl

    A sensitivity analysis utility package for the SciML ecosystem that helps determine how model outputs change relative to input parameters.

  • Project
    pymc-devs/pymc-examples

    PyMC Examples is a repository of tutorials and example notebooks that show how to use PyMC, a Python library for Bayesian statistical modeling and probabilistic machine learning. It provides hands-on, runnable demonstrations to help users learn and apply Bayesian methods.