Dynamic Expert Quantization for Scalable Mixture-of-Experts Inference

arXiv:2511.15015v4 Announce Type: replace-cross Abstract: Mixture-of-Experts (MoE) has become a practical architecture for scaling LLM capacity while keeping per-token compute modest, but deploying MoE models on a single, memory-limited GPU remains…

Read the full story at arXiv cs.AI ↗