pith. sign in

arxiv: 1606.01540 · v1 · pith:QYSKLZTXnew · submitted 2016-06-05 · 💻 cs.LG · cs.AI

OpenAI Gym

Greg Brockman , Vicki Cheung , Ludwig Pettersson , Jonas Schneider , John Schulman , Jie Tang , Wojciech Zaremba This is my paper

Pith reviewed 2026-05-11 19:44 UTC · model grok-4.3

classification 💻 cs.LG cs.AI
keywords reinforcement learningbenchmark environmentscommon interfacealgorithm comparisontoolkitsimulation benchmarksresult sharingresearch infrastructure
0
0 comments X

The pith

A toolkit supplies benchmark problems for reinforcement learning through a shared interface along with a website for comparing algorithm results.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents a toolkit that contains a collection of benchmark problems, each exposing the same interface so that reinforcement learning agents can interact with environments in a consistent way. It pairs this with a website where researchers can share results and directly compare how different algorithms perform on those benchmarks. The work explains the toolkit's components and the design choices made during its creation. A reader would care because this structure could let researchers avoid repeatedly building custom test setups and instead focus on improving methods while seeing clear progress across the field.

Core claim

The central claim is that the toolkit, consisting of a growing collection of benchmark problems that expose a common interface and a website for sharing results, supports reinforcement learning research by enabling standardized testing and performance comparisons. The paper details the toolkit's components and the design decisions that shaped the software.

What carries the argument

The common interface that lets any reinforcement learning algorithm interact uniformly with the benchmark environments.

Load-bearing premise

That providing a common interface for environments plus a platform for sharing results will be sufficient to drive progress and fair comparisons in reinforcement learning.

What would settle it

Track whether new reinforcement learning papers begin using the toolkit's environments for evaluation and posting comparable results on the shared website; sustained low adoption would indicate the standardization has not taken hold.

read the original abstract

OpenAI Gym is a toolkit for reinforcement learning research. It includes a growing collection of benchmark problems that expose a common interface, and a website where people can share their results and compare the performance of algorithms. This whitepaper discusses the components of OpenAI Gym and the design decisions that went into the software.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The manuscript is a whitepaper introducing OpenAI Gym as a toolkit for reinforcement learning research. It describes a growing collection of benchmark environments that share a common interface, a website for sharing results to enable comparison of algorithms, and the software components along with the design decisions that shaped the implementation.

Significance. If the described components are delivered as stated, the work provides a standardized, open-source platform that lowers barriers for RL experimentation and supports reproducible benchmarking across the community. The emphasis on a common interface and public result sharing directly addresses fragmentation in RL evaluation practices.

minor comments (2)
  1. The description of the environment interface in the components section would benefit from an explicit listing of the core methods (e.g., reset, step, render) with their signatures to aid immediate implementation by readers.
  2. A brief note on the versioning or release process for the benchmark collection would clarify how new environments are added while maintaining backward compatibility.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of the OpenAI Gym whitepaper and the recommendation to accept. The referee's summary accurately captures the toolkit's purpose, the common interface for environments, the results-sharing website, and the discussion of design decisions.

Circularity Check

0 steps flagged

No circularity: purely descriptive whitepaper with no derivation chain

full rationale

The manuscript is a software whitepaper that describes the OpenAI Gym toolkit, its environments, common interface, and result-sharing website. It contains no equations, no fitted parameters, no predictions, no formal derivations, and no load-bearing claims that reduce to self-referential inputs. The central content is expository documentation of design choices and released code; the reader's noted assumption about real-world representativeness is not used as a premise for any quantitative or derivational result. No self-citations or ansatzes are invoked in a manner that could create circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is a software toolkit description paper containing no mathematical derivations, fitted parameters, background axioms, or postulated entities.

pith-pipeline@v0.9.0 · 5338 in / 994 out tokens · 65104 ms · 2026-05-11T19:44:10.330352+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 60 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Provably Sub-Linear Two-Timescale NeuroEvolution with Online Plasticity

    cs.NE 2026-06 unverdicted novelty 8.0

    NEOL decouples neuroevolution into outer architecture search and inner online weight adaptation, proving sublinear regret under mild conditions and showing empirical gains over pure NEAT on control benchmarks.

  2. The Rise of AI Teammates in Software Engineering (SE) 3.0: How Autonomous Coding Agents Are Reshaping Software Engineering

    cs.SE 2025-07 conditional novelty 8.0

    AIDev is a new open dataset of 456k AI-agent pull requests showing agents submit code faster than humans but with lower acceptance rates and simpler changes.

  3. BEHAVIOR-1K: A Human-Centered, Embodied AI Benchmark with 1,000 Everyday Activities and Realistic Simulation

    cs.RO 2024-03 accept novelty 8.0

    BEHAVIOR-1K introduces a benchmark of 1,000 human everyday activities in realistic simulated scenes together with the OMNIGIBSON physics simulator to evaluate embodied AI.

  4. Decision Transformer: Reinforcement Learning via Sequence Modeling

    cs.LG 2021-06 accept novelty 8.0

    Decision Transformer casts RL as autoregressive sequence modeling conditioned on desired returns, past states and actions, matching or exceeding offline RL baselines on Atari, Gym and Key-to-Door tasks.

  5. Low-power analogue neural networks with trainable nonlinear connections for continuous control

    cs.LG 2026-06 unverdicted novelty 7.0

    Placing trainable nonlinear functions on connections in analogue networks enables efficient representation of smooth continuous targets with hardware transfer at projected 30 microwatt power.

  6. ENPIRE: Agentic Robot Policy Self-Improvement in the Real World

    cs.AI 2026-06 unverdicted novelty 7.0

    ENPIRE supplies four modules (Environment, Policy Improvement, Rollout, Evolution) that turn real-world robot training into an autonomous optimization loop driven by coding agents.

  7. Expected Free Energy-based Planning as Variational Inference

    cs.AI 2026-06 unverdicted novelty 7.0

    EFE-based planning is formulated as variational free energy minimization with epistemic priors, decomposing into expected plan costs plus a complexity term.

  8. What Type of Inference is Active Inference?

    cs.AI 2026-06 unverdicted novelty 7.0

    EFE-based active inference planning is characterized as VFE on an augmented model plus entropy and planning corrections, with a derived message-passing implementation and grid-world validation.

  9. Task-Induced Representational Invariances Depend on Learning Objective in Deep RL

    cs.LG 2026-06 unverdicted novelty 7.0

    In navigation tasks, DQN learns MDP-homomorphism-invariant representations while PPO learns action-symmetric ones despite comparable performance, with effects on transfer and in LLMs.

  10. FedQHD: Closed-Form Function-Space Federated Reinforcement Learning

    cs.LG 2026-05 unverdicted novelty 7.0

    FedQHD achieves closed-form federated Q-learning via hyperdimensional encoders with linear readouts, formalizes the federation gap under heterogeneous encoders, and reports competitive performance on continuous-state ...

  11. Stochastic MeanFlow Policies: One-Step Generative Control with Entropic Mirror Descent

    cs.LG 2026-05 unverdicted novelty 7.0

    SMFP introduces a one-step generative policy class using MeanFlow to map noise to actions, providing a tractable entropy surrogate for unified off-policy mirror descent training that outperforms Gaussian and generativ...

  12. Proximal State Nudging: Reducing Skill Atrophy from AI Assistance

    cs.RO 2026-05 unverdicted novelty 7.0

    Proximal State Nudging (PSN) jointly optimizes skill development and task performance in shared autonomy, outperforming baselines in LunarLander simulation and yielding up to 7x larger unassisted skill gains with 50% ...

  13. ARC-RL: A Reinforcement Learning Playground Inspired by ARC Raiders

    cs.RO 2026-05 unverdicted novelty 7.0

    ARC-RL provides four new MuJoCo continuous-control environments with hexapod and quadruped morphologies inspired by ARC Raiders, a unified multi-component reward without motion capture, CPG expert demonstrators, and e...

  14. Randomized Advantage Transformation (RAT): Computing Natural Policy Gradients via Direct Backpropagation

    cs.LG 2026-05 unverdicted novelty 7.0

    RAT reformulates regularized natural policy gradients as vanilla gradients with a transformed advantage, computed efficiently via randomized block Kaczmarz iterations on on-policy data.

  15. NeuroTrain: Surveying Local Learning Rules for Spiking Neural Networks with an Open Benchmarking Framework

    cs.NE 2026-05 unverdicted novelty 7.0

    A taxonomy of SNN training algorithms is presented with the release of NeuroTrain, an open benchmarking framework for reproducible comparisons across datasets and architectures.

  16. Distributionally Robust Multi-Task Reinforcement Learning via Adaptive Task Sampling

    cs.LG 2026-05 unverdicted novelty 7.0

    DRATS derives a minimax objective from a feasibility formulation of MTRL to adaptively sample tasks with the largest return gaps, leading to better worst-task performance on MetaWorld benchmarks.

  17. Matrix-Space Reinforcement Learning for Reusing Local Transition Geometry

    cs.LG 2026-05 unverdicted novelty 7.0

    MSRL represents trajectory segments as PSD matrices to prove additive composition properties and bootstrap value functions for better transfer, reaching 0.73 AUC versus 0.57-0.65 baselines.

  18. IGT-OMD: Implicit Gradient Transport for Decision-Focused Learning under Delayed Feedback

    cs.LG 2026-05 unverdicted novelty 7.0

    IGT-OMD reduces gradient transport error from quadratic to linear in delay length for delayed bilevel optimization and achieves sublinear regret with adaptive steps.

  19. gym-invmgmt: An Open Benchmarking Framework for Inventory Management Methods

    cs.LG 2026-05 unverdicted novelty 7.0

    gym-invmgmt is a new benchmarking framework that evaluates inventory policies across optimization and learning methods, finding stochastic programming strongest among non-oracle approaches and PPO-Transformer best amo...

  20. Revisiting Mixture Policies in Entropy-Regularized Actor-Critic

    cs.LG 2026-05 unverdicted novelty 7.0

    A new marginalized reparameterization estimator allows low-variance training of mixture policies in entropy-regularized actor-critic algorithms, matching or exceeding Gaussian policy performance in several continuous ...

  21. Operator-Guided Invariance Learning for Continuous Reinforcement Learning

    cs.LG 2026-05 unverdicted novelty 7.0

    VPSD-RL discovers exact and approximate value-preserving Lie-group operators in continuous RL to stabilize learning via transition augmentation and consistency regularization.

  22. FLUID: Continuous-Time Hyperconnected Sparse Transformer for Sink-Free Learning

    cs.LG 2026-05 unverdicted novelty 7.0

    FLUID is a continuous-time transformer using Liquid Attention Networks to model attention as stable ODE solutions that interpolate between discrete SDPA and CT-RNNs, with an explicit sink gate and liquid hyper-connect...

  23. Learning to Theorize the World from Observation

    cs.LG 2026-05 unverdicted novelty 7.0

    NEO is a probabilistic neural model that induces compositional programs as a learned Language of Thought from non-textual observations and executes them via a shared transition model to enable explanation-driven gener...

  24. EO-Gym: A Multimodal, Interactive Environment for Earth Observation Agents

    cs.AI 2026-05 unverdicted novelty 7.0

    EO-Gym supplies an executable multimodal environment and 9k-trajectory benchmark that turns Earth Observation into a tool-using, multi-step reasoning task, revealing that current VLMs struggle on temporal and cross-se...

  25. Co-Evolving LLM Decision and Skill Bank Agents for Long-Horizon Tasks

    cs.AI 2026-04 unverdicted novelty 7.0

    COSPLAY co-evolves an LLM decision agent with a skill bank agent to improve long-horizon game performance, reporting over 25.1% average reward gains versus frontier LLM baselines on single-player benchmarks.

  26. Hierarchical Active Inference using Successor Representations

    cs.LG 2026-04 unverdicted novelty 7.0

    A hierarchical active inference framework using successor representations learns abstract states and actions to enable efficient planning on navigation and reinforcement learning tasks.

  27. Flow Gym: A framework for the development, benchmarking, training, and deployment of flow-field quantification methods

    physics.flu-dyn 2025-12 accept novelty 7.0

    Flow Gym supplies a JAX-based framework with standardized interfaces, modular components, and utilities to develop, benchmark, train, and deploy flow-field quantification methods such as PIV on both synthetic and expe...

  28. Adaptive Ensemble Aggregation for Actor-Critics

    cs.LG 2025-07 unverdicted novelty 7.0

    AEA dynamically aggregates ensembles in off-policy actor-critics from training dynamics, with proofs of convergence to an error-minimizing equilibrium, bias shrinkage with ensemble size, and monotonic policy improvement.

  29. Steering Your Diffusion Policy with Latent Space Reinforcement Learning

    cs.RO 2025-06 unverdicted novelty 7.0

    DSRL steers pretrained diffusion policies for robotics by applying RL to their latent noise inputs, achieving sample-efficient real-world adaptation with only black-box access.

  30. Group-in-Group Policy Optimization for LLM Agent Training

    cs.LG 2025-05 unverdicted novelty 7.0

    GiGPO adds a hierarchical grouping mechanism to group-based RL so that LLM agents receive both global trajectory and local step-level credit signals, yielding >12% gains on ALFWorld and >9% on WebShop over GRPO while ...

  31. A Generalist Agent

    cs.AI 2022-05 accept novelty 7.0

    Gato is a multi-modal, multi-task, multi-embodiment generalist policy using one transformer network to handle text, vision, games, and robotics tasks.

  32. Mitigating Value Hallucination in Dyna Planning via Multistep Predecessor Models

    cs.LG 2020-06 unverdicted novelty 7.0

    Introduces multistep predecessor models for Dyna planning to mitigate value hallucination by avoiding real-state updates from simulated values.

  33. Dota 2 with Large Scale Deep Reinforcement Learning

    cs.LG 2019-12 accept novelty 7.0

    OpenAI Five achieved superhuman performance in Dota 2 by defeating the world champions using scaled self-play reinforcement learning.

  34. Benchmarking Model-Based Reinforcement Learning

    cs.LG 2019-07 accept novelty 7.0

    Introduces a benchmark suite of over 18 MBRL environments, evaluates multiple algorithms under consistent settings, and identifies three core challenges: dynamics bottleneck, planning horizon dilemma, and early-termin...

  35. Learning the Arrow of Time

    cs.LG 2019-07 unverdicted novelty 7.0

    Introduces a learned arrow of time in MDPs that aligns with the Jordan-Kinderlehrer-Otto notion for stochastic processes and enables practical RL utilities like reachability and side-effect detection.

  36. Exploring Model-based Planning with Policy Networks

    cs.LG 2019-06 unverdicted novelty 7.0

    POPLIN combines policy networks with model-predictive planning by optimizing either action sequences or policy parameters, yielding 3x better sample efficiency than PETS, TD3 and SAC on MuJoCo locomotion tasks.

  37. Soft Actor-Critic Algorithms and Applications

    cs.LG 2018-12 unverdicted novelty 7.0

    SAC extends maximum-entropy RL into a stable off-policy actor-critic method with constrained temperature tuning, outperforming prior algorithms in sample efficiency and consistency on locomotion and manipulation tasks.

  38. Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor

    cs.LG 2018-01 accept novelty 7.0

    Soft Actor-Critic is an off-policy maximum-entropy actor-critic algorithm that achieves state-of-the-art performance and high stability on continuous control benchmarks.

  39. Proximal Policy Optimization Algorithms

    cs.LG 2017-07 accept novelty 7.0

    A clipped surrogate objective L^CLIP = E[min(r_t A_t, clip(r_t, 1-ε, 1+ε) A_t)] enables multi-epoch minibatch policy updates with TRPO-like stability but first-order optimization.

  40. Deep reinforcement learning from human preferences

    stat.ML 2017-06 accept novelty 7.0

    Reinforcement learning agents solve complex tasks without access to the reward function by training a reward predictor from human comparisons of trajectory segments, requiring feedback on less than 1% of interactions.

  41. Visually Grounded Self-Reflection for Vision-Language Models via Reinforcement Learning

    cs.CL 2026-07 unverdicted novelty 6.0

    VRRL trains LVLMs for visually grounded self-reflection via prefix masking and buffered roll-ins, yielding higher out-of-distribution accuracy on grounding and navigation tasks than standard RL baselines.

  42. Dual-Flow Reinforcement Learning with State-Aware Exploration

    cs.LG 2026-06 unverdicted novelty 6.0

    Dual-Flow RL jointly models return distributions and multimodal policies via conditional flow matching with an added ECER for exploration, claiming SOTA results on control benchmarks.

  43. Mosaic: A Benchmark Suite for Differentiable Physics Solvers

    physics.comp-ph 2026-06 unverdicted novelty 6.0

    Mosaic is a benchmark suite evaluating 14 differentiable PDE solvers across fluids, structures, and heat transfer, showing large variations in cost and conditioning but similar convergence behavior.

  44. Solving Markov Decision Processes with Future Information via MPC

    eess.SY 2026-06 unverdicted novelty 6.0

    Parameterized MPC exactly represents optimal policies for MDPs with future information under identified structural conditions and parameters can be learned via RL.

  45. Mass Conservation as an Inductive Bias for Self-Organized Criticality in NCA Reservoirs

    cs.NE 2026-06 unverdicted novelty 6.0

    Mass conservation acts as an inductive bias that promotes stronger self-organized criticality in evolved NCA reservoirs while maintaining comparable performance on sequential memory, classification, and control tasks.

  46. MetaPS: Adaptive Programmatic Strategy Selection for Market Agents

    cs.AI 2026-06 unverdicted novelty 6.0

    MetaPS trains models via simulation rollouts to select from programmatic strategy libraries for market agents, yielding better performance than fixed or direct LLM baselines across model sizes.

  47. Learning the ARTS of Search for Automated Discovery

    cs.AI 2026-06 unverdicted novelty 6.0

    ARTS improves automated scientific discovery by using reasoning LMs with test-time training to separate hypothesis merit from execution quality in tree search, achieving 15.3% relative gains on 22 MLGym and MLEBench tasks.

  48. CRAX: Fast Safe Reinforcement Learning Benchmarking

    cs.LG 2026-06 unverdicted novelty 6.0

    CRAX is a new fast benchmark suite for constrained RL built on MJX, with six environment suites and tasks across difficulty levels, showing no single safe RL method dominates and benefits from curriculum learning.

  49. SVoT: State-aware Visualization-of-Thought for Spatial Reasoning via Reinforcement Learning

    cs.AI 2026-06 unverdicted novelty 6.0

    SVoT uses RL with GRPO to train MLLMs on interleaved textual and visual reasoning chains for multi-hop spatial tasks, achieving up to 65% accuracy gains on new domains with quantitative state verification.

  50. Bilinear gating of motor primitives: a principle linking dendritic computation to rapid goal-directed adaptation

    q-bio.NC 2026-06 unverdicted novelty 6.0

    Burst fraction in motor cortex neurons encodes goal direction via dendritic bilinear gating from apical-basal coincidence, supporting zero-shot generalization in RL agents.

  51. SIMPLE: Simulation-Based Policy Learning and Evaluation for Humanoid Loco-manipulation

    cs.RO 2026-06 unverdicted novelty 6.0

    SIMPLE is a new large-scale simulation benchmark for humanoid loco-manipulation that integrates accurate dynamics and photorealistic rendering and demonstrates policy transfer from simulation to physical robots.

  52. QnRL: Quantum-Native Reinforcement Learning

    quant-ph 2026-06 unverdicted novelty 6.0

    QnRL is a distributional quantum RL framework that distills conditional action policies from moments of quantum generative models in Hilbert space via the QuAK algorithm, reporting higher scores and fewer parameters t...

  53. Closing the Sim-to-Real Gap: An Evaluation Framework for Autonomous Cyber Defense Configuration of Commercial EDR

    cs.CR 2026-06 unverdicted novelty 6.0

    Introduces an evaluation framework for autonomous defense agents hardening commercial EDR, tested in a GOAD lab with Microsoft Defender XDR and two LLMs, revealing three lessons on telemetry design, per-policy attribu...

  54. When Does Deep RL Beat Calibrated Baselines? A Benchmark Study on Adaptive Resource Control

    cs.LG 2026-05 unverdicted novelty 6.0

    Benchmark study finds calibrated rule-based controller outperforms six DRL algorithms on cost for adaptive resource control across workloads, with action-space mismatch explaining large differences in constraint violations.

  55. Reflex: Reinforcement Learning with Reflection Symmetry Exploitation in State-Based Continuous Control

    cs.LG 2026-05 unverdicted novelty 6.0

    Reflex formalizes axial and bilateral reflection symmetries and adds symmetry regularization to PPO and SAC, claiming superior performance and sample efficiency on Gym and DMC benchmarks.

  56. Reflex: Reinforcement Learning with Reflection Symmetry Exploitation in State-Based Continuous Control

    cs.LG 2026-05 unverdicted novelty 6.0

    Reflex formalizes axial and bilateral reflection symmetries and adds symmetry regularization to PPO and SAC, reporting better performance and sample efficiency on Gym and DMC benchmarks.

  57. ARC-RL: A Reinforcement Learning Playground Inspired by ARC Raiders

    cs.RO 2026-05 accept novelty 6.0

    ARC-RL is a new suite of four MuJoCo continuous-control environments featuring game-inspired hexapod and quadruped morphologies, a single closed-form multi-component reward function, CPG demonstrators, and empirical c...

  58. DiPRL: Learning Discrete Programmatic Policies via Architecture Entropy Regularization

    cs.LG 2026-05 unverdicted novelty 6.0

    DiPRL trains nearly discrete programmatic policies in RL by adding architecture entropy regularization to gradient-based optimization, avoiding performance collapse from post-hoc discretization.

  59. Ada-Diffuser: Latent-Aware Adaptive Diffusion for Decision-Making

    cs.LG 2026-05 unverdicted novelty 6.0

    Ada-Diffuser is a causal diffusion model that jointly learns observed interaction structure and underlying latent dynamics from minimal observations for adaptive planning and policy learning.

  60. R2R2: Robust Representation for Intensive Experience Reuse via Redundancy Reduction in Self-Predictive Learning

    cs.LG 2026-05 unverdicted novelty 6.0

    R2R2 introduces a non-centered regularization objective for SPL that addresses conflicts with spectral properties, leading to better performance on continuous control tasks at high UTD ratios.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages · cited by 155 Pith papers · 1 internal anchor

  1. [1]

    Dynamic programming and optimal control

    Dimitri P Bertsekas, Dimitri P Bertsekas, Dimitri P Bertsekas, and Dimitri P Bertsekas. Dynamic programming and optimal control. Athena Scientific Belmont, MA, 1995

    work page 1995

  2. [2]

    V . Mnih, K. Kavukcuoglu, D. Silver, A. A. Rusu, J. Veness, M. G. Bellemare, A. Graves, M. Riedmiller, A. K. Fidjeland, G. Ostrovski, S. Petersen, Sadik Beattie, C., Antonoglou A., H. I., King, D. Kumaran, D. Wierstra, S. Legg, and D. Hassabis. Human-level control through deep reinforcement learning. Nature, 518(7540):529–533, 2015

    work page 2015

  3. [3]

    Schulman, S

    J. Schulman, S. Levine, P. Abbeel, M. I. Jordan, and P. Moritz. Trust region policy optimization. In ICML, pages 1889–1897, 2015

    work page 2015

  4. [4]

    Asynchronous Methods for Deep Reinforcement Learning

    V olodymyr Mnih, Adria Puigdomenech Badia, Mehdi Mirza, Alex Graves, Timothy P Lillicrap, Tim Harley, David Silver, and Koray Kavukcuoglu. Asynchronous methods for deep reinforcement learning. arXiv preprint arXiv:1602.01783, 2016

    work page Pith review arXiv 2016

  5. [5]

    M. G. Bellemare, Y . Naddaf, J. Veness, and M. Bowling. The Arcade Learning Environment: An evaluation platform for general agents. J. Artif. Intell. Res., 47:253–279, 2013

    work page 2013

  6. [6]

    Benchmarking deep reinforcement learning for continuous control

    Yan Duan, Xi Chen, Rein Houthooft, John Schulman, and Pieter Abbeel. Benchmarking deep reinforcement learning for continuous control. arXiv preprint arXiv:1604.06778, 2016

    work page internal anchor Pith review arXiv 2016

  7. [7]

    Geramifard, C

    A. Geramifard, C. Dann, R. H. Klein, W. Dabney, and J. P. How. RLPy: A value-function-based reinforcement learning framework for education and research. J. Mach. Learn. Res., 16:1573–1578, 2015

    work page 2015

  8. [8]

    Tanner and A

    B. Tanner and A. White. RL-Glue: Language-independent software for reinforcement-learning experiments. J. Mach. Learn. Res., 10:2133–2136, 2009

    work page 2009

  9. [9]

    Schaul, J

    T. Schaul, J. Bayer, D. Wierstra, Y . Sun, M. Felder, F. Sehnke, T. R¨uckstieß, and J. Schmidhuber. PyBrain. J. Mach. Learn. Res., 11:743–746, 2010

    work page 2010

  10. [10]

    Abeyruwan

    S. Abeyruwan. RLLib: Lightweight standard and on/off policy reinforcement learning library (C++). http://web.cs.miami.edu/home/saminda/rilib.html, 2013

    work page 2013

  11. [11]

    The reinforcement learning competition 2014

    Christos Dimitrakakis, Guangliang Li, and Nikoalos Tziortziotis. The reinforcement learning competition 2014. AI Magazine, 35(3):61–65, 2014

    work page 2014

  12. [12]

    R. S. Sutton and A. G. Barto. Reinforcement Learning: An Introduction . MIT Press, 1998

    work page 1998

  13. [13]

    Pachi: State of the art open source go program

    Petr Baudi ˇs and Jean-loup Gailly. Pachi: State of the art open source go program. In Advances in Computer Games, pages 24–38. Springer, 2011

    work page 2011

  14. [14]

    Mujoco: A physics engine for model-based control

    Emanuel Todorov, Tom Erez, and Yuval Tassa. Mujoco: A physics engine for model-based control. In Intelligent Robots and Systems (IROS), 2012 IEEE/RSJ International Conference on , pages 5026–5033. IEEE, 2012

    work page 2012

  15. [15]

    ViZDoom: A Doom-based AI Research Platform for Visual Reinforcement Learning

    Michał Kempka, Marek Wydmuch, Grzegorz Runc, Jakub Toczek, and Wojciech Ja´skowski. Vizdoom: A doom-based ai research platform for visual reinforcement learning. arXiv preprint arXiv:1605.02097, 2016. 4

    work page Pith review arXiv 2016