LPM uses a dual-network design to compute intrinsic rewards from the change in prediction error across iterations, providing a noise-robust signal that is theoretically linked to information gain.
hub
Minigrid & mini- world: Modular & customizable reinforcement learning environments for goal-oriented tasks
19 Pith papers cite this work, alongside 3 external citations. Polarity classification is still indexing.
19
Pith papers citing it
3
external citations · Pith
abstract
We present the Minigrid and Miniworld libraries which provide a suite of goal-oriented 2D and 3D environments. The libraries were explicitly created with a minimalistic design paradigm to allow users to rapidly develop new environments for a wide range of research-specific needs. As a result, both have received widescale adoption by the RL community, facilitating research in a wide range of areas. In this paper, we outline the design philosophy, environment details, and their world generation API. We also showcase the additional capabilities brought by the unified API between Minigrid and Miniworld through case studies on transfer learning (for both RL agents and humans) between the different observation spaces. The source code of Minigrid and Miniworld can be found at https://github.com/Farama-Foundation/{Minigrid, Miniworld} along with their documentation at https://{minigrid, miniworld}.farama.org/.
hub tools
citation-role summary
dataset 2
citation-polarity summary
roles
dataset 2representative citing papers
Decoupled surprise-based hierarchical state-space models with top-down inconsistency chunking sustain accurate 250-step open-loop video prediction where prior HSSMs collapse within 10 steps.
Seven disorder-like phenotypes in PPO agents emerge from single appraisal/reward knobs, organize into a two-dimensional affective space, and split into remitting vs treatment-resistant classes.
EvoPolicyGym is a new benchmark suite of 16 compact RL environments that evaluates autonomous policy evolution, with GPT-5.5 achieving the top aggregate rank and top-two performance on all tasks.
A VAE-based latent task representation enables automatic curriculum generation in CRL for non-Euclidean navigation tasks, outperforming interpolation and GAN-based methods in experiments.
DECHRL models causal structures and stochastic delay distributions within hierarchical RL and incorporates them into a delay-aware empowerment objective to improve performance under temporal uncertainty.
H-PPO-Product and H-PPO-SymLoss achieve faster learning and higher final returns than standard PPO and Reward Machine baselines on OfficeWorld, WaterWorld, and DoorKey by transferring imperfect logical policy specifications from easier to harder instances.
Translates Temporal Behavior Trees with LTL constraints into Reward-Petri-Nets that assign structure-based rewards for RL, shown to enable learning where standard methods fail.
This survey categorizes agentic environments for LLMs by eight attributes and domains, introduces symbolic and neural synthesis paradigms with evaluation, and outlines four agent evolution pathways plus three environment evolution paradigms.
An ASP-based implementation of CARCASS abstractions is created and evaluated for RL on two domains.
A C++ Dec-POMDP simulator using data-oriented design and zero-copy PyTorch integration achieves up to 33 million steps per second on a 16-core CPU, enabling multi-agent policy training in minutes with PPO, DQN, and SAC.
Contextual multi-task RL for underwater navigation uses just 1.5% of network weights for task differentiation, mostly from context-variable connections to the first hidden layer.
A reinforcement learning framework extends reward machines with signal temporal logic formulas to generate events and guide training toward satisfying complex task requirements.
A neuro-symbolic DRL approach transfers partial policies as logical rules to bias exploration and rescale Q-values, showing improved performance over reward machine baselines in gridworld environments.
Introduces polychromic objectives adapted into PPO via vine sampling and modified advantages, showing higher success rates and better coverage under perturbations on BabyAI, Minigrid, and algorithmic tasks.
Introduces the Adversarial Rate metric and associated tools to systematically evaluate and visualize the impact of adversarial inputs on DRL policies using formal verification.
ULPS integrates A*-generated symbolic trajectories, fine-tuned BERT priors, MC dropout uncertainty, and entropy-based blending into PPO, reporting over 9% accuracy gains and better sample efficiency on the MiniGridUnlockPickup benchmark.
Synaptic consolidation applied to multi-timescale successor features yields better performance than plasticity-focused methods in RL under gradual environmental drift.
Themis is an XAI-enabled framework for RL from human feedback that supports 200+ environments and includes a scalable cloud platform for collecting human preferences.
citing papers explorer
-
Beyond Noisy-TVs: Noise-Robust Exploration Via Learning Progress Monitoring
LPM uses a dual-network design to compute intrinsic rewards from the change in prediction error across iterations, providing a noise-robust signal that is theoretically linked to information gain.
-
SUNTA: Hierarchical Video Prediction with Surprise-based Chunking
Decoupled surprise-based hierarchical state-space models with top-down inconsistency chunking sustain accurate 250-step open-loop video prediction where prior HSSMs collapse within 10 steps.
-
A Transdiagnostic Space of Disorder Like Phenotypes in Reinforcement Learning Agents
Seven disorder-like phenotypes in PPO agents emerge from single appraisal/reward knobs, organize into a two-dimensional affective space, and split into remitting vs treatment-resistant classes.
-
EvoPolicyGym: Evaluating Autonomous Policy Evolution in Interactive Environments
EvoPolicyGym is a new benchmark suite of 16 compact RL environments that evaluates autonomous policy evolution, with GPT-5.5 achieving the top aggregate rank and top-two performance on all tasks.
-
Curriculum reinforcement learning with measurable task representation learning
A VAE-based latent task representation enables automatic curriculum generation in CRL for non-Euclidean navigation tasks, outperforming interpolation and GAN-based methods in experiments.
-
Delay-Empowered Causal Hierarchical Reinforcement Learning
DECHRL models causal structures and stochastic delay distributions within hierarchical RL and incorporates them into a delay-aware empowerment objective to improve performance under temporal uncertainty.
-
Sample-efficient Neuro-symbolic Proximal Policy Optimization
H-PPO-Product and H-PPO-SymLoss achieve faster learning and higher final returns than standard PPO and Reward Machine baselines on OfficeWorld, WaterWorld, and DoorKey by transferring imperfect logical policy specifications from easier to harder instances.
-
A Reward-Petri-Net Interpretation of Temporal Behavior Trees
Translates Temporal Behavior Trees with LTL constraints into Reward-Petri-Nets that assign structure-based rewards for RL, shown to enable learning where standard methods fail.
-
Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application
This survey categorizes agentic environments for LLMs by eight attributes and domains, introduces symbolic and neural synthesis paradigms with evaluation, and outlines four agent evolution pathways plus three environment evolution paradigms.
-
Answer-Set-Programming-based Abstractions for Reinforcement Learning
An ASP-based implementation of CARCASS abstractions is created and evaluated for RL on two domains.
-
A High-Throughput Compute-Efficient POMDP Hide-And-Seek-Engine (HASE) for Multi-Agent Operations
A C++ Dec-POMDP simulator using data-oriented design and zero-copy PyTorch integration achieves up to 33 million steps per second on a 16-core CPU, enabling multi-agent policy training in minutes with PPO, DQN, and SAC.
-
Task-specific Subnetwork Discovery in Reinforcement Learning for Autonomous Underwater Navigation
Contextual multi-task RL for underwater navigation uses just 1.5% of network weights for task differentiation, mostly from context-variable connections to the first hidden layer.
-
On Tackling Complex Tasks with Reward Machines and Signal Temporal Logics
A reinforcement learning framework extends reward machines with signal temporal logic formulas to generate events and guide training toward satisfying complex task requirements.
-
Sample-Efficient Neurosymbolic Deep Reinforcement Learning
A neuro-symbolic DRL approach transfers partial policies as logical rules to bias exploration and rescale Q-values, showing improved performance over reward machine baselines in gridworld environments.
-
Polychromic Objectives for Reinforcement Learning
Introduces polychromic objectives adapted into PPO via vine sampling and modified advantages, showing higher success rates and better coverage under perturbations on BabyAI, Minigrid, and algorithmic tasks.
-
Analyzing Adversarial Inputs in Deep Reinforcement Learning
Introduces the Adversarial Rate metric and associated tools to systematically evaluate and visualize the impact of adversarial inputs on DRL policies using formal verification.
-
Uncertainty-Aware LLM-Guided Policy Shaping for Sparse-Reward Reinforcement Learning
ULPS integrates A*-generated symbolic trajectories, fine-tuned BERT priors, MC dropout uncertainty, and entropy-based blending into PPO, reporting over 9% accuracy gains and better sample efficiency on the MiniGridUnlockPickup benchmark.
-
Balancing Plasticity and Stability with Fast and Slow Successor Features
Synaptic consolidation applied to multi-timescale successor features yields better performance than plasticity-focused methods in RL under gradual environmental drift.
-
Themis: An explainable AI-enabled framework for Reinforcement Learning with Human Feedback
Themis is an XAI-enabled framework for RL from human feedback that supports 200+ environments and includes a scalable cloud platform for collecting human preferences.