Self-Speculative Decoding for On-device MoE Acceleration.

@inproceedings{zheng-2026-selfspeculative, author = {Peirong Zheng and Wenchao Xu and Haozhao Wang}, editor = {Hakim Hacid and Yoelle Maarek and Francesco Bonchi and Ido Guy and Emine Yilmaz}, title = {{Self-Speculative Decoding for On-device MoE Acceleration}}, 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 = {5155--5164}, publisher = {ACM}, year = {2026}, url = {https://doi.org/10.1145/3774904.3792218}, doi = {https://doi.org/10.1145/3774904.3792218} }