Two Memory Reinforcement

Two-memory reinforcement learning (2MRL) aims to improve the efficiency and performance of reinforcement learning agents by integrating distinct memory systems. Current research focuses on developing architectures that combine the rapid learning of episodic memory with the generalization capabilities of parametric methods like deep reinforcement learning, often employing adaptive memory management and intrinsic motivation to guide exploration. This approach shows promise in addressing challenges like sparse rewards and inefficient search in various applications, from combinatorial optimization to multimodal information retrieval, by enabling more efficient learning and better solution discovery. The resulting improvements in data efficiency and performance have significant implications for the scalability and applicability of reinforcement learning across diverse domains.

Papers

October 6, 2024

AdaMemento: Adaptive Memory-Assisted Policy Optimization for Reinforcement Learning
Renye Yan, Yaozhong Gan, You Wu, Junliang Xing, Ling Liangn, Yeshang Zhu, Yimao Cai
Reinforcement Learning Adaptive Importance Policy OpTimization Memory Mechanism Two Memory Reinforcement

August 5, 2024

MARCO: A Memory-Augmented Reinforcement Framework for Combinatorial Optimization
Andoni I. Garmendia, Quentin Cappart, Josu Ceberio, Alexander Mendiburu
Combinatorial Optimization Problem Neural Combinatorial Optimization Binary Forward Exploration Two Memory Reinforcement

July 17, 2024

MERLIN: Multimodal Embedding Refinement via LLM-based Iterative Navigation for Text-Video Retrieval-Rerank Pipeline
Donghoon Han, Eunhwan Park, Gisang Lee, Adam Lee, Nojun Kwak
Multimodal Phenomenon Video Text Video Text Retrieval Multimodal Retrieval Multimedia Retrieval Retrieve and Rerank Query Embeddings Two Memory Reinforcement

April 20, 2023

Two-Memory Reinforcement Learning
Zhao Yang, Thomas. M. Moerland, Mike Preuss, Aske Plaat
Episodic Memory Two Memory Reinforcement

March 3, 2023

POPGym: Benchmarking Partially Observable Reinforcement Learning
Steven Morad, Ryan Kortvelesy, Matteo Bettini, Stephan Liwicki, Amanda Prorok
Reinforcement Learning Partial Observability Memory Model Two Memory Reinforcement

August 6, 2022

Learning Human Cognitive Appraisal Through Reinforcement Memory Unit
Yaosi Hu, Zhenzhong Chen
Memory Augmentation Memory Mechanism Negative Learning Two Memory Reinforcement

March 24, 2022

MERLIN -- Malware Evasion with Reinforcement LearnINg
Tony Quertier, Benjamin Marais, Stéphane Morucci, Bertrand Fournel
Reinforcement Learning Machine Learning Malware Detection Gradient Based Attack Two Memory Reinforcement