Based Reset

Based resets, a technique involving periodically restarting or resetting parts of an agent's state or environment, are being actively investigated to improve the efficiency and robustness of various learning algorithms. Current research focuses on optimizing reset strategies using reinforcement learning, particularly employing multi-armed bandits and other adaptive methods to dynamically determine when and how to reset, and exploring the impact of different reset types (e.g., partial vs. complete) on exploration and performance. This research is significant for enhancing the sample efficiency and safety of deep reinforcement learning, improving the performance of solvers in computationally hard problems like SAT, and enabling more autonomous learning in robotics and other complex domains.

Papers

October 26, 2024

Self-Normalized Resets for Plasticity in Continual Learning
Vivek F. Farias, Adam D. Jozefiak
Neural Network Continual LEArning Continual Learning Plasticity Rule Global Reset Feature Plasticity Loss Based Reset

April 4, 2024

A Reinforcement Learning based Reset Policy for CDCL SAT Solvers
Chunxiao Li, Charlie Liu, Jonathan Chung, Zhengyang Lu, Piyush Jha, Vijay Ganesh
Reinforcement Learning Automatic Solver Global Reset Feature Based Reset Conflict Driven Clause Learning

January 27, 2024

R$\times$R: Rapid eXploration for Reinforcement Learning via Sampling-based Reset Distributions and Imitation Pre-training
Gagan Khandate, Tristan L. Saidi, Siqi Shang, Eric T. Chang, Yang Liu, Seth Dennis, Johnson Adams, Matei Ciocarlie
Reinforcement Learning Model Free Reinforcement Learning Dexterous Manipulation Based Imitation Rapid Exploration Dexterous in Hand Manipulation Based Reset

October 31, 2023

Sample-Efficient and Safe Deep Reinforcement Learning via Reset Deep Ensemble Agents
Woojun Kim, Yongjae Shin, Jongeui Park, Youngchul Sung
Deep Reinforcement Learning Safe Reinforcement Learning Regret Minimization Deep Reinforcement Learning Agent Based Reset

July 10, 2023

InfLoR-SNN: Reducing Information Loss for Spiking Neural Networks
Yufei Guo, Yuanpei Chen, Liwen Zhang, Xiaode Liu, Xinyi Tong, Yuanyuan Ou, Xuhui Huang, Zhe Ma
Neural Network Spiking Neural Network Binary Spike Information Loss Membrane Potential Based Reset

May 13, 2023

Quantization in Spiking Neural Networks
Bernhard A. Moser, Michael Lunglmayr
Spiking Neural Network Quantization Operator Neuron Model Spike Train Leaky Integrate and Fire Based Reset

January 23, 2023

Congested Bandits: Optimal Routing via Short-term Resets
Pranjal Awasthi, Kush Bhatia, Sreenivas Gollapudi, Kostas Kollias
Multi Armed Bandit Linear Contextual Bandit Linear Reward Based Reset

July 11, 2022

Learning Closed-loop Dough Manipulation Using a Differentiable Reset Module
Carl Qi, Xingyu Lin, David Held
Trajectory Optimization Deformable Object Manipulation Based Reset Initial Trajectory

May 13, 2022

Upside-Down Reinforcement Learning Can Diverge in Stochastic Environments With Episodic Resets
Miroslav Štrupl, Francesco Faccio, Dylan R. Ashley, Jürgen Schmidhuber, Rupesh Kumar Srivastava
Optimal Policy Stochastic Environment Based Reset

April 5, 2022

Automating Reinforcement Learning with Example-based Resets
Jigang Kim, J. hyeon Park, Daesol Cho, H. Jin Kim
Reinforcement Learning Deep Reinforcement Learning Global Reset Feature Based Reset