AMLGym: benchmarking action model learning.

Overview

  • Getting started
  • Passive Algorithms
    • SAM
    • OffLAM
    • NOLAM
    • ROSAME
  • Active Algorithms
    • Random Agent
    • Information Gain Agent
    • OLAM
  • Metrics
    • Syntactic similarity
    • Problem solving
    • Predictive power

Tutorials

  • Learning Domain Models
    • 1. Benchmark domains and trajectories
    • 2. Passive Algorithms
    • 3. Active Algorithms
  • Evaluating Domain Models
    • 1. Syntactic similarity metrics
    • 2. Problem solving metrics
    • 3. Predictive power metrics
  • Tutorial AAAI 2026
  • Tutorial AAMAS 2026

API

  • amlgym
    • amlgym package
      • amlgym.algorithms package
      • amlgym.benchmarks package
      • amlgym.metrics package
      • amlgym.modeling package
      • amlgym.util package
AMLGym: benchmarking action model learning.
  • Overview: module code

All modules for which code is available

  • amlgym.algorithms
    • amlgym.algorithms.ActiveAlgorithmAdapter
    • amlgym.algorithms.InformationGainAgent
    • amlgym.algorithms.NOLAM
    • amlgym.algorithms.OLAM
    • amlgym.algorithms.OffLAM
    • amlgym.algorithms.PassiveAlgorithmAdapter
    • amlgym.algorithms.ROSAME
    • amlgym.algorithms.RandomAgent
    • amlgym.algorithms.SAM
  • amlgym.benchmarks
  • amlgym.metrics
    • amlgym.metrics._predictive
    • amlgym.metrics._solving
    • amlgym.metrics._syntactic
  • amlgym.modeling.UPEnv
  • amlgym.modeling.env
  • amlgym.modeling.trajectory
  • amlgym.util.SimpleDomainReader
  • amlgym.util.gen_problems
  • amlgym.util.gen_states_predictability
  • amlgym.util.gen_trajs_learning
  • amlgym.util.gen_trajs_learning_hard
  • amlgym.util.gen_trajs_predictability
  • amlgym.util.util

© Copyright 2025, Roni Stern, Leonardo Lamanna, Argaman Mordoch, Yarin Benyamin, Pascal Lauer, Brendan Juba, Gregor Behnke, Christian Muise, Pascal Bercher, Mauro Vallati, Kai Xi, Omar Wattad, Omer Eliyahu.

Built with Sphinx using a theme provided by Read the Docs.