Zobrazeno 1 - 10
of 858
pro vyhledávání: '"Levine, Sergey"'
Autor:
Kang, Katie, Setlur, Amrith, Ghosh, Dibya, Steinhardt, Jacob, Tomlin, Claire, Levine, Sergey, Kumar, Aviral
Despite the remarkable capabilities of modern large language models (LLMs), the mechanisms behind their problem-solving abilities remain elusive. In this work, we aim to better understand how the learning dynamics of LLM finetuning shapes downstream
Externí odkaz:
http://arxiv.org/abs/2411.07681
Recent progress on large language models (LLMs) has enabled dialogue agents to generate highly naturalistic and plausible text. However, current LLM language generation focuses on responding accurately to questions and requests with a single effectiv
Externí odkaz:
http://arxiv.org/abs/2411.05194
Value-based reinforcement learning (RL) can in principle learn effective policies for a wide range of multi-turn problems, from games to dialogue to robotic control, including via offline RL from static previously collected datasets. However, despite
Externí odkaz:
http://arxiv.org/abs/2411.05193
Assistive agents should make humans' lives easier. Classically, such assistance is studied through the lens of inverse reinforcement learning, where an assistive agent (e.g., a chatbot, a robot) infers a human's intention and then selects actions to
Externí odkaz:
http://arxiv.org/abs/2411.02623
Autor:
Black, Kevin, Brown, Noah, Driess, Danny, Esmail, Adnan, Equi, Michael, Finn, Chelsea, Fusai, Niccolo, Groom, Lachy, Hausman, Karol, Ichter, Brian, Jakubczak, Szymon, Jones, Tim, Ke, Liyiming, Levine, Sergey, Li-Bell, Adrian, Mothukuri, Mohith, Nair, Suraj, Pertsch, Karl, Shi, Lucy Xiaoyang, Tanner, James, Vuong, Quan, Walling, Anna, Wang, Haohuan, Zhilinsky, Ury
Robot learning holds tremendous promise to unlock the full potential of flexible, general, and dexterous robot systems, as well as to address some of the deepest questions in artificial intelligence. However, bringing robot learning to the level of g
Externí odkaz:
http://arxiv.org/abs/2410.24164
Reinforcement learning (RL) holds great promise for enabling autonomous acquisition of complex robotic manipulation skills, but realizing this potential in real-world settings has been challenging. We present a human-in-the-loop vision-based RL syste
Externí odkaz:
http://arxiv.org/abs/2410.21845
Offline goal-conditioned reinforcement learning (GCRL) is a major problem in reinforcement learning (RL) because it provides a simple, unsupervised, and domain-agnostic way to acquire diverse behaviors and representations from unlabeled data without
Externí odkaz:
http://arxiv.org/abs/2410.20092
Autor:
Hatch, Kyle B., Balakrishna, Ashwin, Mees, Oier, Nair, Suraj, Park, Seohong, Wulfe, Blake, Itkina, Masha, Eysenbach, Benjamin, Levine, Sergey, Kollar, Thomas, Burchfiel, Benjamin
Image and video generative models that are pre-trained on Internet-scale data can greatly increase the generalization capacity of robot learning systems. These models can function as high-level planners, generating intermediate subgoals for low-level
Externí odkaz:
http://arxiv.org/abs/2410.20018
Sample-efficient online reinforcement learning often uses replay buffers to store experience for reuse when updating the value function. However, uniform replay is inefficient, since certain classes of transitions can be more relevant to learning. Wh
Externí odkaz:
http://arxiv.org/abs/2410.18082
Unsupervised pretraining has been transformative in many supervised domains. However, applying such ideas to reinforcement learning (RL) presents a unique challenge in that fine-tuning does not involve mimicking task-specific data, but rather explori
Externí odkaz:
http://arxiv.org/abs/2410.18076