Zobrazeno 1 - 10
of 886
pro vyhledávání: '"Li, Yi Chen"'
Low sample efficiency is an enduring challenge of reinforcement learning (RL). With the advent of versatile large language models (LLMs), recent works impart common-sense knowledge to accelerate policy learning for RL processes. However, we note that
Externí odkaz:
http://arxiv.org/abs/2407.03964
Autor:
Li, Yi-Chen, Zhang, Fuxiang, Qiu, Wenjie, Yuan, Lei, Jia, Chengxing, Zhang, Zongzhang, Yu, Yang, An, Bo
Large Language Models (LLMs), trained on a large amount of corpus, have demonstrated remarkable abilities. However, it may not be sufficient to directly apply open-source LLMs like Llama to certain real-world scenarios, since most of them are trained
Externí odkaz:
http://arxiv.org/abs/2407.03856
Large language models (LLMs) have catalyzed a paradigm shift in natural language processing, yet their limited controllability poses a significant challenge for downstream applications. We aim to address this by drawing inspiration from the neural me
Externí odkaz:
http://arxiv.org/abs/2405.17039
Autor:
Lin, Haoxin, Xu, Yu-Yan, Sun, Yihao, Zhang, Zhilong, Li, Yi-Chen, Jia, Chengxing, Ye, Junyin, Zhang, Jiaji, Yu, Yang
Model-based methods in reinforcement learning offer a promising approach to enhance data efficiency by facilitating policy exploration within a dynamics model. However, accurately predicting sequential steps in the dynamics model remains a challenge
Externí odkaz:
http://arxiv.org/abs/2405.17031
Autor:
Jia, Chengxing, Zhang, Fuxiang, Li, Yi-Chen, Gao, Chen-Xiao, Liu, Xu-Hui, Yuan, Lei, Zhang, Zongzhang, Yu, Yang
Offline meta-reinforcement learning (OMRL) proficiently allows an agent to tackle novel tasks while solely relying on a static dataset. For precise and efficient task identification, existing OMRL research suggests learning separate task representati
Externí odkaz:
http://arxiv.org/abs/2403.07261
Autor:
Zhang, Xinyu, Qiu, Wenjie, Li, Yi-Chen, Yuan, Lei, Jia, Chengxing, Zhang, Zongzhang, Yu, Yang
Developing policies that can adjust to non-stationary environments is essential for real-world reinforcement learning applications. However, learning such adaptable policies in offline settings, with only a limited set of pre-collected trajectories,
Externí odkaz:
http://arxiv.org/abs/2402.11317
Autor:
Chen, Xiong-Hui, Ye, Junyin, Zhao, Hang, Li, Yi-Chen, Shi, Haoran, Xu, Yu-Yan, Ye, Zhihao, Yang, Si-Hang, Huang, Anqi, Xu, Kai, Zhang, Zongzhang, Yu, Yang
Imitation learning (IL) enables agents to mimic expert behaviors. Most previous IL techniques focus on precisely imitating one policy through mass demonstrations. However, in many applications, what humans require is the ability to perform various ta
Externí odkaz:
http://arxiv.org/abs/2310.05712
We consider the problem of learning the best possible policy from a fixed dataset, known as offline Reinforcement Learning (RL). A common taxonomy of existing offline RL works is policy regularization, which typically constrains the learned policy by
Externí odkaz:
http://arxiv.org/abs/2306.06569
Autor:
Li, Yi Chen1,2 (AUTHOR), Chen, Yung Hao2 (AUTHOR), Chang, Shen Chang3 (AUTHOR), Lin, Min Jung4 (AUTHOR), Lin, Li Jen5 (AUTHOR), Lee, Tzu Tai1,2,6,7 (AUTHOR) ttlee@dragon.nchu.edu.tw
Publikováno v:
Italian Journal of Animal Science. Dec2024, Vol. 23 Issue 1, p1091-1103. 13p.
