paper-with-me

홈 › Papers

Working Memory Graphs

2019-11-17 · ICML 2020 1 · Ricky Loynd, Roland Fernandez, Asli Celikyilmaz, Adith Swaminathan, Matthew Hausknecht

Transformers have increasingly outperformed gated RNNs in obtaining new state-of-the-art results on supervised tasks involving text sequences. Inspired by this trend, we study the question of how Transformer-based models can improve the performance of sequential decision-making agents. We present the Working Memory Graph (WMG), an agent that employs multi-head self-attention to reason over a dynamic set of vectors representing observed and recurrent state. We evaluate WMG in three environments featuring factored observation spaces: a Pathfinding environment that requires complex reasoning over past observations, BabyAI gridworld levels that involve variable goals, and Sokoban which emphasizes future planning. We find that the combination of WMG's Transformer-based architecture with factored observation spaces leads to significant gains in learning efficiency compared to baseline architectures across all tasks. WMG demonstrates how Transformer-based models can dramatically boost sample efficiency in RL environments for which observations can be factored.

📄 PDF Abstract BibTeX arXiv:1911.07141

Code (0)

등록된 구현이 없습니다.

Tasks

Decision MakingSequential Decision MakingSokoban

Similar Papers 제목 키워드 기반

Bayesian Functional Connectivity and Graph Convolutional Network for Working Memory Load Classification

2024-04-30 · Harshini Gangapuram, Vidya Manian

Brain responses related to working memory originate from distinct brain areas and oscillate at different frequencies. EEG signals with high temporal correlation can effectively capture these responses. Therefore, estimat…

ClassificationConnectivity EstimationEEGFunctional Connectivity

Learning with Memory Embeddings

2015-11-25 · Volker Tresp, Cristóbal Esteban, Yinchong Yang, Stephan Baier 외

Embedding learning, a.k.a. representation learning, has been shown to be able to model large-scale semantic knowledge graphs. A key concept is a mapping of the knowledge graph to a tensor representation whose entries are…

Knowledge GraphsRepresentation Learning

GraphCogent: Mitigating LLMs' Working Memory Constraints via Multi-Agent Collaboration in Complex Graph Understanding

2025-08-17 · Rongzheng Wang, Shuang Liang, Qizhi Chen, Yihong Huang 외 arxiv

Large language models (LLMs) show promising performance on small-scale graph reasoning tasks but fail when handling real-world graphs with complex queries. This phenomenon arises from LLMs' working memory constraints, wh…

MemSifter: Offloading LLM Memory Retrieval via Outcome-Driven Proxy Reasoning

2026-03-03 · Jiejun Tan, Zhicheng Dou, Liancheng Zhang, Yuyang Hu 외 arxiv

As Large Language Models (LLMs) are increasingly used for long-duration tasks, maintaining effective long-term memory has become a critical challenge. Current methods often face a trade-off between cost and accuracy. Sim…

Reinforcement Learning

AWM: Answerable Working Memory for Long-Document VQA Agents

2026-08-26 · Dongzhuoran Zhou, Yuqicheng Zhu, Yule Liu, Zhen Yang 외 arxiv

Long-document visual question answering increasingly relies on VLM agents that retrieve candidate pages, inspect page images, write findings to working memory, and synthesize answers. Working memory should carry answer-s…

Visual Question Answering