Knowledge Transfer in Model-Based Reinforcement Learning Agents for Efficient Multi-Task Learning

Simple item page
dc.contributor.authorKuzmenko, Dmytro en_US
dc.contributor.authorShvai, Nadiya en_US
dc.date.accessioned2025-11-19T06:47:04Z
dc.date.available2025-11-19T06:47:04Z
dc.date.issued2025
dc.description.abstractWe propose an efficient knowledge transfer approach for modelbased reinforcement learning, addressing the challenge of deploying large world models in resource-constrained environments. Our method distills a high-capacity multi-task agent (317M parameters) into a compact 1M parameter model, achieving state-of-the-art performance on the MT30 benchmark with a normalized score of 28.45, a substantial improvement over the original 1M parameter model’s score of 18.93. This demonstrates the ability of our distillation technique to consolidate complex multi-task knowledge effectively. Additionally, we apply FP16 post-training quantization, reducing the model size by 50% while maintaining performance. Our work bridges the gap between the power of large models and practical deployment constraints, offering a scalable solution for efficient and accessible multi-task reinforcement learning in robotics and other resource-limited domains. en_US
dc.identifier.citationKuzmenko D. Knowledge Transfer in Model-Based Reinforcement Learning Agents for Efficient Multi-Task Learning / Dmytro Kuzmenko, Nadiya Shvai // Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS. - 2025. - P. 2597-2599. - https://doi.org/10.48550/arXiv.2501.05329 en_US
dc.identifier.urihttps://doi.org/10.48550/arXiv.2501.05329
dc.identifier.urihttps://ekmair.ukma.edu.ua/handle/123456789/37605
dc.language.isoen en_US
dc.relation.sourceProceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, AAMAS en_US
dc.statusfirst published en_US
dc.subjectModel-Based Reinforcement Learning en_US
dc.subjectMulti-Task Learning en_US
dc.subjectKnowledge Distillation en_US
dc.subjectModel Compression en_US
dc.subjectEfficient RL Agents en_US
dc.subjectconference materials en_US
dc.titleKnowledge Transfer in Model-Based Reinforcement Learning Agents for Efficient Multi-Task Learning en_US
dc.typeConference materials en_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Kuzmenko_Shvai_Knowledge_Transfer_in_Model-Based_Reinforcement_Learning_Agents_for_Efficient_Multi-Task_Learning.pdf
Size:
923.22 KB
Format:
Adobe Portable Document Format
Download
Collections
Кафедра математики
Кафедра мультимедійних систем