MoEcho: Exploiting Side-Channel Attacks to Compromise User Privacy in Mixture-of

Abstract

Abstract: The transformer architecture has become a cornerstone of modern AI, fueling remarkable progress across applications in natural language processing, computer vision, and multi-modal learning. As these models continue to scale explosively for performance, implementation efficiency remains a critical challenge. Mixture-of-Experts (MoE) architectures, selectively activating specialized subnetworks (experts), offer a unique balance between model accuracy and computational cost. However, the adaptive routing in MoE architectures—where input tokens are dynamically directed to specialized experts base

Research Question

How does the accuracy of SMoES-based MoE-VLMs with soft modality-guided routing compare to dense models of equivalent parameter count on the MMMU benchmark across 7B to 34B scales, and what is the performance gap trend?

Verification Level

Descriptive public verification status only; aggregate claim counts are public, but individual claim records are not exposed here.

Quality Tier

Descriptive public triage only; this tier does not alter current publication or DOI behavior.

Quality Dimensions

Automated triage signals derived from public fields; not human peer review or independent validation.

Correction Record

Public corrections are additive records. Current status does not claim the synthesis is error-free.

Provenance