VisMMoE exploits visual-expert affinity via token pruning to achieve up to 2.68x faster VL-MoE inference on memory-constrained hardware while keeping accuracy competitive.
LLM -Based Edge Intelligence: A Comprehensive Survey on Architectures, Applications, Security and Trustworthiness
3 Pith papers cite this work, alongside 138 external citations. Polarity classification is still indexing.
3
Pith papers citing it
138
external citations · external index
citation-role summary
background 1
citation-polarity summary
years
2026 3roles
background 1polarities
background 1representative citing papers
SocaSim uses LLM-based multi-agent simulations to model Putnam's Social Capital Theory, reproducing macro-level patterns and aligning with human group-level decisions, then applies the framework to smart elderly care adoption.
Human-in-the-loop LLM system for emerging topic detection in tax feedback shows better expert alignment than baselines via similarity analysis and officer review.
citing papers explorer
-
VisMMOE: Exploiting Visual-Expert Affinity for Efficient Visual-Language MoE Offloading
VisMMoE exploits visual-expert affinity via token pruning to achieve up to 2.68x faster VL-MoE inference on memory-constrained hardware while keeping accuracy competitive.
-
From Blueprint to Reality: Modeling and Applying Putnam's Social Capital Theory with LLM-based Multi-agent Simulations
SocaSim uses LLM-based multi-agent simulations to model Putnam's Social Capital Theory, reproducing macro-level patterns and aligning with human group-level decisions, then applies the framework to smart elderly care adoption.
-
LLM-based Models for Detecting Emerging Topics in Service Feedback
Human-in-the-loop LLM system for emerging topic detection in tax feedback shows better expert alignment than baselines via similarity analysis and officer review.