AI RESEARCH
Rethinking Model Efficiency: Multi-Agent Inference with Large Models
arXiv CS.CV
•
ArXi:2604.04929v1 Announce Type: new Most vision-language models (VLMs) apply a large language model (LLM) as the decoder, where the response tokens are generated sequentially through autoregression. Therefore, the number of output tokens can be the bottleneck of the end-to-end latency. However, different models may require vastly different numbers of output tokens to achieve comparable performance. In this work, we conduct a comprehensive analysis of the latency across different components of VLMs on simulated data.