04:00
2026-10-08
arxiv.org
artificial-intelligence
MoR-MLLM: Mixture of Recursions for Efficient Multimodal Large Language Models
Researchers proposed MoR-MLLM, a computation-sparse multimodal large language model built on the Mixture-of-Recursions (MoR) framework that uses adaptive per-token recursion to dynamically adjust recu…