Visual token pruning in MLLMs fails on complex reasoning due to Relevant Visual Information Shift during decoding, but the DSTP framework fixes it training-free across models.
Llavolta: Efficient multi-modal models via stage-wise visual context compression
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 6roles
background 1polarities
background 1representative citing papers
PyramidDrop accelerates LVLMs by staged, similarity-based dropping of visual tokens that become redundant in deeper layers, delivering 40% faster training and 55% lower inference cost with comparable accuracy.
VideoChat-Flash applies hierarchical video token compression to achieve ~50x reduction in context length for long videos while maintaining near-original performance on long-context benchmarks.
TASM proposes a task-aware structured memory framework using task-vector compression, bipartite token merging, and a Core Memory plus Latent Bank hierarchy to enable efficient dynamic multi-modal in-context learning.
TwigVLM adds a twig module to VLMs for twig-guided token pruning and self-speculative decoding, retaining 96% performance after pruning 88.9% visual tokens and delivering 154% speedup on long responses for LLaVA-1.5-7B.
InternVideo3 introduces Multimodal Contextual Reasoning and M^2LA attention to enable closed-loop evidence accumulation in long-video understanding and agentic tool use, reporting strong benchmark results.
citing papers explorer
-
Why and When Visual Token Pruning Fails? A Study on Relevant Visual Information Shift in MLLMs Decoding
Visual token pruning in MLLMs fails on complex reasoning due to Relevant Visual Information Shift during decoding, but the DSTP framework fixes it training-free across models.
-
PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction
PyramidDrop accelerates LVLMs by staged, similarity-based dropping of visual tokens that become redundant in deeper layers, delivering 40% faster training and 55% lower inference cost with comparable accuracy.
-
VideoChat-Flash: Hierarchical Compression for Long-Context Video Modeling
VideoChat-Flash applies hierarchical video token compression to achieve ~50x reduction in context length for long videos while maintaining near-original performance on long-context benchmarks.
-
Task-Aware Structured Memory for Dynamic Multi-modal In-Context Learning
TASM proposes a task-aware structured memory framework using task-vector compression, bipartite token merging, and a Core Memory plus Latent Bank hierarchy to enable efficient dynamic multi-modal in-context learning.
-
Growing a Multi-head Twig via Distillation and Reinforcement Learning to Accelerate Large Vision-Language Models
TwigVLM adds a twig module to VLMs for twig-guided token pruning and self-speculative decoding, retaining 96% performance after pruning 88.9% visual tokens and delivering 154% speedup on long responses for LLaVA-1.5-7B.
-
InternVideo3: Agentify Foundation Models with Multimodal Contextual Reasoning
InternVideo3 introduces Multimodal Contextual Reasoning and M^2LA attention to enable closed-loop evidence accumulation in long-video understanding and agentic tool use, reporting strong benchmark results.