paper-with-me

홈 › Papers

Grow Your Limits: Continuous Improvement with Real-World RL for Robotic Locomotion

2023-10-26 · Laura Smith, YunHao Cao, Sergey Levine

Deep reinforcement learning (RL) can enable robots to autonomously acquire complex behaviors, such as legged locomotion. However, RL in the real world is complicated by constraints on efficiency, safety, and overall training stability, which limits its practical applicability. We present APRL, a policy regularization framework that modulates the robot's exploration over the course of training, striking a balance between flexible improvement potential and focused, efficient exploration. APRL enables a quadrupedal robot to efficiently learn to walk entirely in the real world within minutes and continue to improve with more training where prior work saturates in performance. We demonstrate that continued training with APRL results in a policy that is substantially more capable of navigating challenging situations and is able to adapt to changes in dynamics with continued training.

📄 PDF Abstract BibTeX arXiv:2310.17634

Code (0)

등록된 구현이 없습니다.

Tasks

Deep Reinforcement LearningEfficient ExplorationReinforcement Learning (RL)

Similar Papers 제목 키워드 기반

Muizalix Pro

2025-02-07 · 02/07 2025 2 · Muizalix Pro

At Muizalix, we’re a dedicated team of creative professionals and digital experts passionate about helping businesses grow. Specializing in social media marketing, SEO, content creation, and website design, we provide co…

Marketing

Continuous-Time Deep Glioma Growth Models

2021-06-23 · Jens Petersen, Fabian Isensee, Gregor Köhler, Paul F. Jäger 외

The ability to estimate how a tumor might evolve in the future could have tremendous clinical benefits, from improved treatment decisions to better dose distribution in radiation therapy. Recent work has approached the g…

Time SeriesTime Series AnalysisVariational Inference

Learning Perturbations to Extrapolate Your LLM

2026-05-13 · Zetai Cen, Chenfei Gu, Jin Zhu, Ting Li 외 arxiv

Recent advancements in large language models demonstrate that injecting perturbations can substantially enhance extrapolation performance. However, current approaches often rely on discrete perturbations with fixed desig…

When Your Own Output Becomes Your Training Data: Noise-to-Meaning Loops and a Formal RSI Trigger

2025-05-05 · Rintaro Ando

We present Noise-to-Meaning Recursive Self-Improvement (N2M-RSI), a minimal formal model showing that once an AI agent feeds its own outputs back as inputs and crosses an explicit information-integration threshold, its i…

AI AgentAutoML

Trading Strategies with Position Limits

2017-12-19

Whether you trade futures for yourself or a hedge fund, your strategy is counted. Long and short position limits make the number of unique strategies finite. Formulas of the numbers of strategies, transactions, do nothin…

Position