Agnostic Exploration

Agnostic exploration in reinforcement learning focuses on developing agents capable of efficiently exploring environments without prior knowledge of specific tasks or reward functions. Current research emphasizes improving sample efficiency through methods like causal exploration, leveraging structured world models to guide exploration, and incorporating intrinsic motivation such as curiosity or entropy maximization. These advancements aim to enable more robust and adaptable agents, impacting fields like robotics, where efficient learning in unstructured environments is crucial, and improving the generalization capabilities of machine learning models across diverse tasks.

Papers

July 30, 2024

Boosting Efficiency in Task-Agnostic Exploration through Causal Knowledge
Yupei Yang, Biwei Huang, Shikui Tu, Lei Xu
Causal Knowledge Enhancing Efficiency Causal Insight Agnostic Exploration Agnostic Reinforcement Learning Causal Exploration

July 15, 2024

Last-Iterate Global Convergence of Policy Gradients for Constrained Reinforcement Learning
Alessandro Montenegro, Marco Mussi, Matteo Papini, Alberto Maria Metelli
Policy Gradient Primal Dual Constrained Reinforcement Learning Last Iterate Convergence Stochastic Policy Agnostic Exploration

July 9, 2024

A Unified Approach to Multi-task Legged Navigation: Temporal Logic Meets Reinforcement Learning
Jesse Jiang, Samuel Coogan, Ye Zhao
Reinforcement Learning Unified Framework Legged Robot Temporal Logic Robot Motion Planning Goal Directed Behavior Agnostic Exploration Exploration Task Hopping Robot

January 8, 2024

Curiosity & Entropy Driven Unsupervised RL in Multiple Environments
Shaurya Dewan, Anisha Jain, Zoe LaLena, Lifan Yu
Entropy Regularized Unsupervised Reinforcement Learning Curiosity Inducing Situation Curiosity Driven Exploration Alpha MEPOL Agnostic Exploration

October 11, 2023

NoMaD: Goal Masked Diffusion Policies for Navigation and Exploration
Ajay Sridhar, Dhruv Shah, Catherine Glossop, Sergey Levine
Environment Exploration Local Navigation Navigation Task Diffusion Policy Agnostic Exploration

June 21, 2023

Investigating Poor Performance Regions of Black Boxes: LIME-based Exploration in Sepsis Detection
Mozhgan Salimiparsa, Surajsinh Parmar, San Lee, Choongmin Kim, Yonghwan Kim, Jang Yong Kim
Inherent Interpretability Black Box Local Interpretable Model Agnostic Explanation Performance Degradation Agnostic Exploration Sepsis Detection Sepsis Diagnosis

May 17, 2023

Reward-agnostic Fine-tuning: Provable Statistical Benefits of Hybrid Reinforcement Learning
Gen Li, Wenhao Zhan, Jason D. Lee, Yuejie Chi, Yuxin Chen
Reinforcement Learning Model Based Offline Reinforcement Learning Provable Benefit Hybrid Reinforcement Learning Reward Fine Tuning Agnostic Exploration Policy Fine Tuning

April 14, 2023

Minimax-Optimal Reward-Agnostic Exploration in Reinforcement Learning
Gen Li, Yuling Yan, Yuxin Chen, Jianqing Fan
Reinforcement Learning Reward Free Exploration Agnostic Exploration Sample Efficient Offline Reinforcement Learning

September 9, 2022

Task-Agnostic Learning to Accomplish New Tasks
Xianqi Zhang, Xingtao Wang, Xu Liu, Wenrui Wang, Xiaopeng Fan, Debin Zhao
Reinforcement Learning Imitation Learning New Task Task Agnostic Agnostic Exploration Sequential Attachment Based Fragment Embedding

June 22, 2022

Curious Exploration via Structured World Models Yields Zero-Shot Object Manipulation
Cansu Sancaktar, Sebastian Blaes, Georg Martius
Intrinsic Reward Object Manipulation Guided Exploration Agnostic Exploration Intelligent Exploration

November 25, 2021

Interesting Object, Curious Agent: Learning Task-Agnostic Exploration
Simone Parisi, Victoria Dean, Deepak Pathak, Abhinav Gupta
Arbitrary Object Exploration Method Binary Forward Exploration Agnostic Exploration