Soft Actor Critic

Soft Actor-Critic (SAC) is a deep reinforcement learning algorithm aiming to learn robust and efficient policies by maximizing both expected reward and policy entropy. Current research focuses on improving SAC's sample efficiency, addressing safety constraints through methods like Lagrangian formulations and meta-gradient optimization, and extending its applicability to various domains including robotics, autonomous driving, and multi-agent systems. These advancements are significant because they enhance the practicality and reliability of reinforcement learning for real-world applications requiring safe and efficient decision-making in complex environments.

Papers

October 26, 2024

Velocity-History-Based Soft Actor-Critic Tackling IROS'24 Competition "AI Olympics with RealAIGym"
Tim Lukas Faust, Habib Maraqten, Erfan Aghadavoodi, Boris Belousov, Jan Peters
Competition Platform Soft Actor Critic Underactuated System Entropy Regularized AI City Challenge

October 24, 2024

Learn 2 Rage: Experiencing The Emotional Roller Coaster That Is Reinforcement Learning
Lachlan Mares, Stefan Podgorski, Ian Reid
Reinforcement Learning Autonomous Driving Autonomous Racing Soft Actor Critic Emotional Stimulus

October 22, 2024

Corrected Soft Actor Critic for Continuous Control
Yanjun Chen, Xinming Zhang, Xianghui Wang, Zhiqiang Xu, Xiaoyu Shen, Wei Zhang
Deep Reinforcement Learning Continuous Control Faster Convergence Soft Actor Critic Action Sampling Soft Actor

September 29, 2024

Constrained Reinforcement Learning for Safe Heat Pump Control
Baohe Zhang, Lilli Frison, Thomas Brox, Joschka Bödecker
Constrained Optimization Constrained Reinforcement Learning Soft Actor Critic Thermal Management Building Energy

September 8, 2024

Soft Actor-Critic with Beta Policy via Implicit Reparameterization Gradients
Luca Della Libera
Deep Reinforcement Learning Actor Critic Algorithm Soft Actor Critic Reparameterization Method Frank Wolfe Policy Optimization

August 15, 2024

Meta SAC-Lag: Towards Deployable Safe Reinforcement Learning via MetaGradient-based Hyperparameter Tuning
Homayoun Honari, Amir Mehdi Soufi Enayati, Mehran Ghafarian Tamizi, Homayoun Najjaran
Safe Reinforcement Learning Hyperparameter Tuning Safe Exploration Soft Actor Critic Time Delay

August 4, 2024

RVI-SAC: Average Reward Off-Policy Deep Reinforcement Learning
Yukinari Hisaki, Isao Ono
Soft Actor Critic Policy Deep Reinforcement Learning Average Reward Average Reward Reinforcement Learning

May 2, 2024

S$^2$AC: Energy-Based Reinforcement Learning with Stein Soft Actor Critic
Safa Messaoud, Billel Mokeddem, Zhenghai Xue, Linsey Pang, Bo An, Haipeng Chen, Sanjay Chawla
Soft Actor Critic Energy Based Stein Variational Gradient Descent Maximum Entropy Reinforcement Learning Expressive Policy MaxEnt RL

April 4, 2024

AdaGlimpse: Active Visual Exploration with Arbitrary Glimpse Position and Scale
Adam Pardyl, Michał Wronka, Maciej Wołczyk, Kamil Adamczewski, Tomasz Trzciński, Bartosz Zieliński
Visual Analogue Scale Reinforcement Learning Algorithm Soft Actor Critic Visual Exploration Exploration Task

April 3, 2024

Grid-Mapping Pseudo-Count Constraint for Offline Reinforcement Learning
Yi Shen, Hanyan Huang
Reinforcement Learning Offline Reinforcement Learning Soft Actor Critic

April 2, 2024

Learning from Demonstration Framework for Multi-Robot Systems Using Interaction Keypoints and Soft Actor-Critic Methods
Vishnunandan L. N. Venkatesh, Byung-Cheol Min
Reinforcement Learning LeArning Abstract Multi Robot System Multi Robot Task Soft Actor Critic Demonstration Framework Robot Object Interaction Contact Aware Interaction Segmentation

March 21, 2024

Constrained Reinforcement Learning with Smoothed Log Barrier Function
Baohe Zhang, Yuan Zhang, Lilli Frison, Thomas Brox, Joschka Bödecker
Reinforcement Learning Policy Learning Reward Shaping Constrained Reinforcement Learning Soft Actor Critic Control Constraint Logarithmic Barrier

March 18, 2024

Decomposing Control Lyapunov Functions for Efficient Reinforcement Learning
Antonio Lopez, David Fridovich-Keil
Reinforcement Learning Efficient Reinforcement Learning Soft Actor Critic Hamilton Jacobi Reachability Control Lyapunov Function

March 1, 2024

Overestimation, Overfitting, and Plasticity in Actor-Critic: the Bitter Lesson of Reinforcement Learning
Michal Nauman, Michał Bortkiewicz, Piotr Miłoś, Tomasz Trzciński, Mateusz Ostaszewski, Marek Cygan
Reinforcement Learning Regularization Model Policy Reinforcement Learning Model Overfitting Regularization Technique Plasticity Rule Soft Actor Critic Policy Algorithm

February 24, 2024

Discretionary Lane-Change Decision and Control via Parameterized Soft Actor-Critic for Hybrid Action Space
Yuan Lin, Xiao Liu, Zishun Zheng
Autonomous Driving External Control Action Space Lane Change Soft Actor Critic

February 15, 2024

Risk-Sensitive Soft Actor-Critic for Robust Deep Reinforcement Learning under Distribution Shifts
Tobias Enders, James Harrison, Maximilian Schiffer
Distribution Shift Risk Sensitive Robust Policy Soft Actor Critic Risk Sensitive Reinforcement Learning Robust Deep Reinforcement Learning Risk Sensitive Algorithm

January 31, 2024

Nash Soft Actor-Critic LEO Satellite Handover Management Algorithm for Flying Vehicles
Jinxuan Chen, Mustafa Ozger, Cicek Cavdar
Multi Agent Reinforcement Learning Robot Handover Soft Actor Critic Low Earth Orbit Satellite

October 23, 2023

Enhancing Robotic Manipulation: Harnessing the Power of Multi-Task Reinforcement Learning and Single Life Reinforcement Learning in Meta-World
Ghadi Nehme, Ishan Sabane, Tejas Y. Deo
Real Power Proximal Policy Optimization Robotic Manipulation Multi Task Reinforcement Learning Soft Actor Critic Single Agent Reinforcement Learning Trust Region Policy Optimization Possible World

September 7, 2023

Hybrid of representation learning and reinforcement learning for dynamic and complex robotic motion planning
Chengmin Zhou, Xin Lu, Jiapeng Dai, Bingding Huang, Xiaoxu Liu, Pasi Fränti
Reinforcement Learning Representation Learning Motion Planning Efficient Hybrid Soft Actor Critic Robot Motion Planning Based Planner

September 1, 2023

Parallel Distributional Prioritized Deep Reinforcement Learning for Unmanned Aerial Vehicles
Alisson Henrique Kolling, Victor Augusto Kich, Junior Costa de Jesus, Andressa Cavalcante da Silva, Ricardo Bedin Grando, Paulo Lilles Jorge Drews-Jr, Daniel F. T. Gamarra
Deep Reinforcement Learning Unmanned Aerial Vehicle Soft Actor Critic Autonomous Aerial Navigation Distributional Policy Gradient

Soft Actor Critic

Papers

Velocity-History-Based Soft Actor-Critic Tackling IROS'24 Competition "AI Olympics with RealAIGym"

Learn 2 Rage: Experiencing The Emotional Roller Coaster That Is Reinforcement Learning

Corrected Soft Actor Critic for Continuous Control

Constrained Reinforcement Learning for Safe Heat Pump Control

Soft Actor-Critic with Beta Policy via Implicit Reparameterization Gradients

Meta SAC-Lag: Towards Deployable Safe Reinforcement Learning via MetaGradient-based Hyperparameter Tuning

RVI-SAC: Average Reward Off-Policy Deep Reinforcement Learning

S$^2$AC: Energy-Based Reinforcement Learning with Stein Soft Actor Critic

AdaGlimpse: Active Visual Exploration with Arbitrary Glimpse Position and Scale

Grid-Mapping Pseudo-Count Constraint for Offline Reinforcement Learning

Learning from Demonstration Framework for Multi-Robot Systems Using Interaction Keypoints and Soft Actor-Critic Methods

Constrained Reinforcement Learning with Smoothed Log Barrier Function

Decomposing Control Lyapunov Functions for Efficient Reinforcement Learning

Overestimation, Overfitting, and Plasticity in Actor-Critic: the Bitter Lesson of Reinforcement Learning

Discretionary Lane-Change Decision and Control via Parameterized Soft Actor-Critic for Hybrid Action Space

Risk-Sensitive Soft Actor-Critic for Robust Deep Reinforcement Learning under Distribution Shifts

Nash Soft Actor-Critic LEO Satellite Handover Management Algorithm for Flying Vehicles

Enhancing Robotic Manipulation: Harnessing the Power of Multi-Task Reinforcement Learning and Single Life Reinforcement Learning in Meta-World

Hybrid of representation learning and reinforcement learning for dynamic and complex robotic motion planning

Parallel Distributional Prioritized Deep Reinforcement Learning for Unmanned Aerial Vehicles