Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
Open Access
- 27 October 2025
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 5110-5119
Abstract
No abstract availableAll Related Versions
Funding Information
- The National Natural Science Foundation of China ((Grant No. 62276245))
This publication has 13 references indexed in Scilit:
- Omni-SMoLA: Boosting Generalist Multimodal Models with Soft Mixture of Low-Rank Experts2024
- HRVDA: High-Resolution Visual Document Assistant2024
- VILA: On Pre-training for Visual Language Models2024
- A Survey of Multimodel Large Language Models2024
- Attention Where It Matters: Rethinking Visual Document Understanding with Selective Region Concentration2023
- FlexMoE: Scaling Large-scale Sparse Pre-trained Model Training via Dynamic Device PlacementProceedings of the ACM on Management of Data, 2023
- FasterMoE2022
- GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering2019
- Towards VQA Models That Can Read2019
- VizWiz Grand Challenge: Answering Visual Questions from Blind People2018