: Accelerating Agentic Multimodal LLMs via Speculative Perception and Planning
- 25 September 2026
- book chapter
- Published by Springer Nature
- p. 1-20
Abstract
No abstract availableKeywords
All Related Versions
This publication has 28 references indexed in Scilit:
- MoE-LLaVA: Mixture of Experts for Large Vision-Language ModelsIEEE Transactions on Multimedia, 2026
- HERO: Rethinking Visual Token Early Dropping in High-Resolution Large Vision-Language ModelsIEEE Transactions on Multimedia, 2026
- Feather the Throttle: Revisiting Visual Token Pruning for Vision-Language Model Acceleration2025
- Growing a Twig to Accelerate Large Vision-Language Models2025
- FrameFusion: Combining Similarity and Importance for Video Token Reduction on Large Vision Language Models2025
- A Stitch in Time Saves Nine: Small VLM is a Precise Guidance for Accelerating Large VLMs2025
- VisionZip: Longer is Better but Not Necessary in Vision Language Models2025
- Efficient Inference of Vision Instruction-Following Models with Elastic Cache2024
- V*: Guided Visual Search as a Core Mechanism in Multimodal LLMs2024
- BranchyNet: Fast inference via early exiting from deep neural networks2016