POPGym Arcade: Parallel Pixelated POMDPs

2025-03-03 · Zekang Wang, Zhe He, Borong Zhang, Edan Toledo, Steven Morad

We present the POPGym Arcade, a collection of hardware-accelerated, pixel-based environments with shared observation and action spaces. Each environment includes fully and partially observable variants, enabling counterfactual studies on partial observability. We also introduce mathematical tools for analyzing policies under partial observability, which reveal how agents recall past information to make decisions. Our analysis shows (1) that controlling for partial observability is critical and (2) that agents with long-term memory learn brittle policies that struggle to generalize. Finally, we demonstrate that recurrent policies can be "poisoned" by old, out-of-distribution observations, with implications for sim-to-real transfer, imitation learning, and offline reinforcement learning.

📄 PDF Abstract BibTeX arXiv:2503.01450

Code (1)

bolt-research/popgym_arcade 공식 구현 jax

Tasks

counterfactualImitation LearningQ-Learning

Methods 이 논문이 사용한 방법론

Library 설명 없음

POPGym: Benchmarking Partially Observable Reinforcement Learning

2023-03-03 · Steven Morad, Ryan Kortvelesy, Matteo Bettini, Stephan Liwicki 외

Real world applications of Reinforcement Learning (RL) are often partially observable, thus requiring memory. Despite this, partial observability is still largely ignored by contemporary RL benchmarks and libraries. We i…

Benchmarking GPU Partially Observable Reinforcement Learning reinforcement-learning+4

Compact Pixelated Microstrip Forward Broadside Coupler Using Binary Particle Swarm Optimization

2024-08-27 · Kourosh Parsaei, Rasool Keshavarz, Rashid Mirzavand Boroujeni, Negin Shariati

In this paper, a compact microstrip forward broadside coupler (MFBC) with high coupling level is proposed in the frequency band of 3.5-3.8 GHz. The coupler is composed of two parallel pixelated transmission lines. To val…

Pixelated Semantic Colorization

2019-01-27 · Jiaojiao Zhao, Jungong Han, Ling Shao, Cees G. M. Snoek

While many image colorization algorithms have recently shown the capability of producing plausible color versions from gray-scale photographs, they still suffer from limited semantic understanding. To address this shortc…

Colorization Image Colorization Object Semantic Segmentation

Massively Parallel Methods for Deep Reinforcement Learning

2015-07-15 · Arun Nair, Praveen Srinivasan, Sam Blackwell, Cagdas Alcicek 외

We present the first massively distributed architecture for deep reinforcement learning. This architecture uses four main components: parallel actors that generate new behaviour; parallel learners that are trained from s…

Atari Games Deep Reinforcement Learning reinforcement-learning Reinforcement Learning+1

Ultra-Fast and Efficient Design Method Using Deep Learning for Capacitive Coupling WPT System

2024-08-21 · Rasool Keshavarz, Ehsan Majidi, Ali Raza, Negin Shariati

Capacitive coupling wireless power transfer (CCWPT) is one of the pervasive methods to transfer power in the reactive near-field zone. In this paper, a flexible design methodology based on Binary Particle Swarm Optimizat…