Research
Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization
arXiv:2609.03158v1 Announce Type: cross Abstract: Visual token pruning reduces the inference cost of vision-language models (VLMs), but most methods only ask which tokens to keep. This retained-token
arXiv:2609.03158v1 Announce Type: cross Abstract: Visual token pruning reduces the inference cost of vision-language models (VLMs), but most methods only ask which tokens to keep. This retained-token view can keep redundant high-scoring tokens while leaving discarded evidence without a close representative. We propose CoverPruner, a training-free pruner that asks the complementary demand-side question: after a token is removed, which surviving original token represents it for the target VLM? CoverPruner formulates pruning as Representational Coverage Maximization (RCM), covering the full projected visual-token set with query-weighted demand. It instantiates RCM with projector-space coverage and a lightweight first-layer attention probe. Across multiple VLM architectures and compression rates, CoverPruner achieves the best average accuracy among all compared methods, with the largest gains usually appearing under aggressive compression.
Related
- Beyond Surrogate Gradients: Fully Differentiable Token Pruning for Vision-Language Models
- HiPrune: Hierarchical Attention for Efficient Token Pruning in Vision-Language Models
- OccamToken: Efficient VLM Inference with Training-Free and Budget-Adaptive Token Pruning
- Evading Visual Aphasia: Contrastive Adaptive Semantic Token Pruning for Vision-Language Models
Source: arXiv cs.CL | 2026-09-04