Copertina

Preventing Learning Stagnation in PPO by Scaling to 1 Million Parallel Environments

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Beukman, Michael, Khetarpal, Khimya, Zheng, Zeyu, Dabney, Will, Foerster, Jakob, Dennis, Michael, Lyle, Clare
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Disentangling the Causes of Plasticity Loss in Neural Networks
di: Lyle, Clare, et al.
Pubblicazione: (2024)

Normalization and effective learning rates in reinforcement learning
di: Lyle, Clare, et al.
Pubblicazione: (2024)

JaxUED: A simple and useable UED library in Jax
di: Coward, Samuel, et al.
Pubblicazione: (2024)

Robust Intervention Learning from Emergency Stop Interventions
di: Pronovost, Ethan, et al.
Pubblicazione: (2026)

Refining Minimax Regret for Unsupervised Environment Design
di: Beukman, Michael, et al.
Pubblicazione: (2024)

Kinetix: Investigating the Training of General Agents through Open-Ended Physics-Based Control Tasks
di: Matthews, Michael, et al.
Pubblicazione: (2024)

A Unifying Framework for Action-Conditional Self-Predictive Reinforcement Learning
di: Khetarpal, Khimya, et al.
Pubblicazione: (2024)

High entropy leads to symmetry equivariant policies in Dec-POMDPs
di: Forkel, Johannes, et al.
Pubblicazione: (2025)

An Optimisation Framework for Unsupervised Environment Design
di: Monette, Nathan, et al.
Pubblicazione: (2025)

Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement Learning
di: Matthews, Michael, et al.
Pubblicazione: (2024)

Goal-Conditioned Agents that Learn Everything All at Once
di: Matthews, Michael, et al.
Pubblicazione: (2026)

No Regrets: Investigating and Improving Regret Approximations for Curriculum Discovery
di: Rutherford, Alexander, et al.
Pubblicazione: (2024)

Plasticity as the Mirror of Empowerment
di: Abel, David, et al.
Pubblicazione: (2025)

Self-Predictive Representations for Combinatorial Generalization in Behavioral Cloning
di: Lawson, Daniel, et al.
Pubblicazione: (2025)

Optimizing Return Distributions with Distributional Dynamic Programming
di: Pires, Bernardo Ávila, et al.
Pubblicazione: (2025)

Near-Minimax-Optimal Distributional Reinforcement Learning with a Generative Model
di: Rowland, Mark, et al.
Pubblicazione: (2024)

Representation Learning via Non-Contrastive Mutual Information
di: Guo, Zhaohan Daniel, et al.
Pubblicazione: (2025)

Improving Regret Approximation for Unsupervised Dynamic Environment Generation
di: Mead, Harry, et al.
Pubblicazione: (2026)

Discovering Minimal Reinforcement Learning Environments
di: Liesen, Jarek, et al.
Pubblicazione: (2024)

Learning Multi-Agent Communication with Contrastive Learning
di: Lo, Yat Long, et al.
Pubblicazione: (2023)

Affordances Enable Partial World Modeling with LLMs
di: Khetarpal, Khimya, et al.
Pubblicazione: (2026)

Mixtures of Experts Unlock Parameter Scaling for Deep RL
di: Obando-Ceron, Johan, et al.
Pubblicazione: (2024)

JaxLife: An Open-Ended Agentic Simulator
di: Lu, Chris, et al.
Pubblicazione: (2024)

The Yokai Learning Environment: Tracking Beliefs Over Space and Time
di: Ruhdorfer, Constantin, et al.
Pubblicazione: (2025)

Multi-Agent Craftax: Benchmarking Open-Ended Multi-Agent Reinforcement Learning at the Hyperscale
di: Omari, Bassel Al, et al.
Pubblicazione: (2025)

Colored Noise in PPO: Improved Exploration and Performance through Correlated Action Sampling
di: Hollenstein, Jakob, et al.
Pubblicazione: (2023)

Lifelong Reinforcement Learning via Neuromodulation
di: Lee, Sebastian, et al.
Pubblicazione: (2024)

Edge Caching Optimization with PPO and Transfer Learning for Dynamic Environments
di: Niknia, Farnaz, et al.
Pubblicazione: (2024)

TabFlex: Scaling Tabular Learning to Millions with Linear Attention
di: Zeng, Yuchen, et al.
Pubblicazione: (2025)

RobocupGym: A challenging continuous control benchmark in Robocup
di: Beukman, Michael, et al.
Pubblicazione: (2024)

Step-Size Decay and Structural Stagnation in Greedy Sparse Learning
di: Berná, Pablo M.
Pubblicazione: (2026)

Directional-Clamp PPO
di: Karpel, Gilad, et al.
Pubblicazione: (2025)

Mirror Learning: A Unifying Framework of Policy Optimisation
di: Kuba, Jakub Grudzien, et al.
Pubblicazione: (2022)

The Edge-of-Reach Problem in Offline Model-Based Reinforcement Learning
di: Sims, Anya, et al.
Pubblicazione: (2024)

SOReL and TOReL: Two Methods for Fully Offline Reinforcement Learning
di: Fellows, Mattie, et al.
Pubblicazione: (2025)

Context Parallelism for Scalable Million-Token Inference
di: Yang, Amy, et al.
Pubblicazione: (2024)

What Can Grokking Teach Us About Learning Under Nonstationarity?
di: Lyle, Clare, et al.
Pubblicazione: (2025)

Abstraction for Offline Goal-Conditioned Reinforcement Learning
di: Wibault, Clarisse, et al.
Pubblicazione: (2026)

Parallel Scaling Law for Language Models
di: Chen, Mouxiang, et al.
Pubblicazione: (2025)

OPPO: Accelerating PPO-based RLHF via Pipeline Overlap
di: Yan, Kaizhuo, et al.
Pubblicazione: (2025)