GRADE: Personalized Multi-Task Fusion via Group-Relative Reinforcement Learning with Adaptive Dirichlet Exploration.

@inproceedings{hong-2026-grade, author = {Tingfeng Hong and Pingye Ren and Xinlong Xiao and Chao Wang and Chenyi Lei and Wenwu Ou and Han Li}, editor = {Hakim Hacid and Yoelle Maarek and Francesco Bonchi and Ido Guy and Emine Yilmaz}, title = {{GRADE: Personalized Multi-Task Fusion via Group-Relative Reinforcement Learning with Adaptive Dirichlet Exploration}}, booktitle = {{Proceedings of the ACM Web Conference 2026, WWW 2026, Dubai, United Arab Emirates, originally scheduled for April 13-17, 2026, rescheduled for June 29 - July 3, 2026}}, pages = {8232--8240}, publisher = {ACM}, year = {2026}, url = {https://doi.org/10.1145/3774904.3792844}, doi = {https://doi.org/10.1145/3774904.3792844} }