@article{lynxenablingefficientmoeinferencethrough, title = {Lynx: Enabling Efficient MoE Inference through Dynamic Batch-Aware Expert Selection}, author = {Vima Gupta and Kartik Sinha and Ada Gavrilovska and Anand Padmanabha Iyer}, year = {2024}, eprint = {2411.08982}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2411.08982v1}, }