arXiv AI
Don't Send What You Don't Need: Question-Guided Token Pruning as a Privacy Defense for Vision-Language Models
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv AI
Sep 1
Defending Wearable VLMs Against Private Attribute Inference
arXiv:2608.28691v1 Announce Type: cross Abstract: Wearable VLM pipelines promise continuous multimodal assistance from egocentric visual capture: a user asks a task-driven question about the surround...
arXiv Computer Vision
Sep 14
USPLIT-VQA: U-Shaped Split Learning for Visual Question Answering with Contribution-Aware Weighted Aggregation
arXiv:2609.12168v1 Announce Type: new Abstract: Visual Question Answering (VQA) systems, jointly interpreting images and natural language queries, hold significant promise across many domains, yet th...
arXiv Machine Learning
Sep 4
Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization
CoverPruner is a training‑free visual token pruner that reframes token pruning as a representational coverage maximization problem. Instead of selecting which tokens to keep, it asks which surviving token best represents each removed token for a vision‑language model. Experiments on various VLM architectures show that CoverPruner consistently outperforms existing methods, especially under high compression rates.
arXiv Machine Learning
Aug 13
LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection
arXiv:2608. 11691v1 Announce Type: new Abstract: Reinforcement-learning (RL) post-training equips multimodal large reasoning models (MLRMs) with exploratory chains of thought (CoT), substantially improving visual reasoning.
arXiv Computer Vision
3d ago
MiCo: Mutual Information Coverage Optimization through Semantic Erasure Modeling for Efficient MLLM Inference
arXiv:2609.34330v2 Announce Type: replace Abstract: Multimodal large language models (MLLMs) have demonstrated impressive performance in multimodal understanding, but processing large numbers of visu...
arXiv Computer Vision
3d ago
Feature-Aware Token Attack for Compression-Triggered Stealthy Failures in Large Vision-Language Models
arXiv:2609.39134v1 Announce Type: new Abstract: Visual-token compression improves the efficiency of large vision-language models, but can expose failures that full-token evaluation misses. We study a...