AI Agent - arXivDaily 专题

2106.06091 2022-02-14 cs.LG cs.AI 62%

DECORE: Deep Compression with Reinforcement Learning

Manoj Alwani, Yang Wang, Vashisht Madhavan

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2202.03983 2022-02-09 cs.LG cs.AI 62%

Provable Reinforcement Learning with a Short-Term Memory

Yonathan Efroni, Chi Jin, Akshay Krishnamurthy, Sobhan Miryoosefi

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2202.02790 2022-02-08 cs.LG cs.AI 62%

Learning Synthetic Environments and Reward Networks for Reinforcement Learning

Fabio Ferreira, Thomas Nierhoff, Andreas Saelinger, Frank Hutter

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Journal ref International Conference on Learning Representations (ICLR 2022)

2110.03743 2022-02-01 cs.LG cs.AI cs.IT math.IT 62%

Reinforcement Learning in Reward-Mixing MDPs

Jeongyeol Kwon, Yonathan Efroni, Constantine Caramanis, Shie Mannor

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments NeurIPS 2021; fixed typo

2106.03926 2022-02-01 cs.AI cs.LG 62%

Reconciling Rewards with Predictive State Representations

Andrea Baisero, Christopher Amato

专题命中记忆与上下文管理：planning(abstract)；分类 cs.AI、cs.LG

Comments IJCAI 2021

Journal ref Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence (2021) Pages 2170-2176

2102.05599 2022-01-19 cs.LG cs.AI 62%

Improving Model-Based Reinforcement Learning with Internal State Representations through Self-Supervision

Julien Scholz, Cornelius Weber, Muhammad Burhan Hafez, Stefan Wermter

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Journal ref Proc. Intl. Joint Conf. Neural Networks (IJCNN), 2021, forthcoming

2201.03538 2022-01-11 cs.AI cs.LG cs.MA 62%

Assisting Unknown Teammates in Unknown Tasks: Ad Hoc Teamwork under Partial Observability

João G. Ribeiro, Cassandro Martinho, Alberto Sardinha, Francisco S. Melo

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2201.02628 2022-01-11 cs.LG cs.AI 62%

Attention Option-Critic

Raviteja Chunduru, Doina Precup

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2011.12906 2022-01-04 cs.CV cs.AI cs.LG eess.IV eess.SP 62%

A Review of Open-World Learning and Steps Toward Open-World Learning Without Labels

Mohsen Jafarzadeh, Akshay Raj Dhamija, Steve Cruz, Chunchun Li, Touqeer Ahmad, Terrance E. Boult

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2107.07394 2021-12-30 cs.LG cs.AI 62%

Explore and Control with Adversarial Surprise

Arnaud Fickinger, Natasha Jaques, Samyak Parajuli, Michael Chang, Nicholas Rhinehart, Glen Berseth, Stuart Russell, Sergey Levine

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2112.13141 2021-12-28 cs.LG cs.AI cs.NA math.NA 62%

On the Unreasonable Efficiency of State Space Clustering in Personalization Tasks

Anton Dereventsov, Ranga Raju Vatsavai, Clayton Webster

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2109.10736 2021-12-28 cs.LG cs.AI stat.ML 62%

Estimation Error Correction in Deep Reinforcement Learning for Deterministic Actor-Critic Methods

Baturay Saglam, Enes Duran, Dogan C. Cicek, Furkan B. Mutlu, Suleyman S. Kozat

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2107.08408 2021-12-24 cs.CL cs.AI cs.MA cs.RO 62%

Pre-trained Language Models as Prior Knowledge for Playing Text-based Games

Ishika Singh, Gargi Singh, Ashutosh Modi

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.CL

Comments 40 Pages (8 Pages main content + 1 Page references + 31 Pages Appendix). Some new results added

2111.14973 2021-12-23 cs.CV cs.AI cs.LG cs.RO 62%

MultiPath++: Efficient Information Fusion and Trajectory Aggregation for Behavior Prediction

Balakrishnan Varadarajan, Ahmed Hefny, Avikalp Srivastava, Khaled S. Refaat, Nigamaa Nayakanti, Andre Cornman, Kan Chen, Bertrand Douillard, Chi Pang Lam, Dragomir Anguelov, Benjamin Sapp

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2112.06336 2021-12-14 cs.AI cs.LG 62%

Representing Knowledge as Predictions (and State as Knowledge)

Mark Ring

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments Other than a few edits suggested by generous colleagues, this paper has not changed since roughly 2013. Thus, some aspects of it are now dated; for example, GVFs (aka "forecasts") and off-policy learning are now well known. Nevertheless, I believe this paper still has useful insights to offer the community, especially the growing community of enthusiastic researchers in Continual Learning

2112.03899 2021-12-08 cs.LG cs.AI 62%

Information is Power: Intrinsic Control via Information Capture

Nicholas Rhinehart, Jenny Wang, Glen Berseth, John D. Co-Reyes, Danijar Hafner, Chelsea Finn, Sergey Levine

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments NeurIPS 2021

2112.02342 2021-12-07 cs.LG cs.AI cs.NE 62%

Overcome Anterograde Forgetting with Cycled Memory Networks

Jian Peng, Dingqi Ye, Bo Tang, Yinjie Lei, Yu Liu, Haifeng Li

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments 14 pages, 15 figures

2112.01163 2021-12-03 cs.LG cs.AI cs.RO 62%

Robust Robotic Control from Pixels using Contrastive Recurrent State-Space Models

Nitish Srivastava, Walter Talbott, Martin Bertran Lopez, Shuangfei Zhai, Josh Susskind

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments NeurIPS Deep Reinforcement Learning Workshop 2021. Code can be found at https://github.com/apple/ml-core

2111.06389 2021-11-23 cs.RO cs.AI cs.CV cs.LG cs.SY eess.SY 62%

Full-Body Visual Self-Modeling of Robot Morphologies

Boyuan Chen, Robert Kwiatkowski, Carl Vondrick, Hod Lipson

专题命中记忆与上下文管理：planning(abstract)；分类 cs.AI、cs.LG

Comments Project website: https://robot-morphology.cs.columbia.edu/

2106.02757 2021-11-23 cs.LG cs.AI 62%

Heuristic-Guided Reinforcement Learning

Ching-An Cheng, Andrey Kolobov, Adith Swaminathan

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2111.02104 2021-11-09 cs.LG cs.AI 62%

Model-Based Episodic Memory Induces Dynamic Hybrid Controls

Hung Le, Thommen Karimpanal George, Majid Abdolshah, Truyen Tran, Svetha Venkatesh

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments 26 pages

1903.00715 2021-11-03 cs.LG cs.AI stat.ML 62%

Efficient Reinforcement Learning for StarCraft by Abstract Forward Models and Transfer Learning

Ruo-Ze Liu, Haifeng Guo, Xiaozhong Ji, Yang Yu, Zhen-Jia Pang, Zitai Xiao, Yuzhou Wu, Tong Lu

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2111.00987 2021-11-02 econ.GN cs.AI cs.LG cs.MA q-fin.EC 62%

Modelling the transition to a low-carbon energy supply

Alexander Kell

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments PhD thesis

2009.11459 2021-11-02 cs.AI cs.LG cs.RO cs.SY eess.SY math.OC 62%

Robust Finite-State Controllers for Uncertain POMDPs

Murat Cubuktepe, Nils Jansen, Sebastian Junges, Ahmadreza Marandi, Marnix Suilen, Ufuk Topcu

专题命中记忆与上下文管理：planning(abstract)；分类 cs.AI、cs.LG

2107.12931 2021-10-29 cs.LG cs.AI cs.RO 62%

Autonomous Reinforcement Learning via Subgoal Curricula

Archit Sharma, Abhishek Gupta, Sergey Levine, Karol Hausman, Chelsea Finn

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

2110.14555 2021-10-28 cs.LG cs.AI cs.GT cs.MA stat.ML 62%

V-Learning -- A Simple, Efficient, Decentralized Algorithm for Multiagent RL

Chi Jin, Qinghua Liu, Yuanhao Wang, Tiancheng Yu

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments This is the journal version of arXiv:2006.12007, with new results on (1) finding CE and CCE in the multiplayer general-sum setting, (2) monotonic techniques that allow V-learning to output Markov policies in a subset of settings, and (3) decoupling V-learning with the adversarial bandit subroutine

2110.12680 2021-10-26 cs.CL cs.AI 62%

TODSum: Task-Oriented Dialogue Summarization with State Tracking

Lulu Zhao, Fujia Zheng, Keqing He, Weihao Zeng, Yuejie Lei, Huixing Jiang, Wei Wu, Weiran Xu, Jun Guo, Fanyu Meng

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.CL

Comments Preprint

2110.10735 2021-10-26 cs.LG cs.AI 62%

Dynamic Bottleneck for Robust Self-Supervised Exploration

Chenjia Bai, Lingxiao Wang, Lei Han, Animesh Garg, Jianye Hao, Peng Liu, Zhaoran Wang

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.LG

Comments NeurIPS 2021

2110.09796 2021-10-20 cs.LG cs.AI 62%

Offline Reinforcement Learning with Value-based Episodic Memory

Xiaoteng Ma, Yiqin Yang, Hao Hu, Qihan Liu, Jun Yang, Chongjie Zhang, Qianchuan Zhao, Bin Liang

专题命中记忆与上下文管理：planning(abstract)；分类 cs.AI、cs.LG

2008.02742 2021-10-20 cs.CL cs.AI cs.RO 62%

Compositional Networks Enable Systematic Generalization for Grounded Language Understanding

Yen-Ling Kuo, Boris Katz, Andrei Barbu

专题命中记忆与上下文管理：agent(abstract)；分类 cs.AI、cs.CL

Comments Accepted in Findings of EMNLP 2021

1. 记忆与上下文管理 4753 篇

DECORE: Deep Compression with Reinforcement Learning

Provable Reinforcement Learning with a Short-Term Memory

Learning Synthetic Environments and Reward Networks for Reinforcement Learning

Reinforcement Learning in Reward-Mixing MDPs

Reconciling Rewards with Predictive State Representations

Improving Model-Based Reinforcement Learning with Internal State Representations through Self-Supervision

Assisting Unknown Teammates in Unknown Tasks: Ad Hoc Teamwork under Partial Observability

Attention Option-Critic

A Review of Open-World Learning and Steps Toward Open-World Learning Without Labels

Explore and Control with Adversarial Surprise

On the Unreasonable Efficiency of State Space Clustering in Personalization Tasks

Estimation Error Correction in Deep Reinforcement Learning for Deterministic Actor-Critic Methods

Pre-trained Language Models as Prior Knowledge for Playing Text-based Games

MultiPath++: Efficient Information Fusion and Trajectory Aggregation for Behavior Prediction

Representing Knowledge as Predictions (and State as Knowledge)

Information is Power: Intrinsic Control via Information Capture

Overcome Anterograde Forgetting with Cycled Memory Networks

Robust Robotic Control from Pixels using Contrastive Recurrent State-Space Models

Full-Body Visual Self-Modeling of Robot Morphologies

Heuristic-Guided Reinforcement Learning

Model-Based Episodic Memory Induces Dynamic Hybrid Controls

Efficient Reinforcement Learning for StarCraft by Abstract Forward Models and Transfer Learning

Modelling the transition to a low-carbon energy supply

Robust Finite-State Controllers for Uncertain POMDPs

Autonomous Reinforcement Learning via Subgoal Curricula

V-Learning -- A Simple, Efficient, Decentralized Algorithm for Multiagent RL

TODSum: Task-Oriented Dialogue Summarization with State Tracking

Dynamic Bottleneck for Robust Self-Supervised Exploration

Offline Reinforcement Learning with Value-based Episodic Memory

Compositional Networks Enable Systematic Generalization for Grounded Language Understanding