Cut-ViT: Task-Specific Model Pruning via Gram Anchoring Subspace Consistency

arXiv:2608.28205v1 Announce Type: new Abstract: Pruning visual foundation models has attracted considerable attention. However, existing methods focus on rigid point-to-point token alignment on a single dataset for pruning, suffering from two limitations: i) robustness degradation, and ii) task-specificity deficiency. To address these limitations, we propose a task-specific pruning pipeline, named Cut-ViT. Specifically, we first construct gram anchoring matrices from both spatial and semantic perspectives, and perform the subspace decomposition to extract the corresponding subspace bases. Basis-agnostic and residual constraints are then adopted to align the gram subspaces between the native and pruned DINOv3 models along spatial and channel dimensions, enabling subnetworks to inherit robust feature representations of native DINOv3. Furthermore, we design spectral entropy adaptation, which quantifies the information density of feature manifolds along spatial and channel dimensions, thereby adapting the pruning objective to specific downstream tasks. Experiments show that Cut-ViT requires approximately one minute on a single A100 GPU to obtain subnetworks at various sparsity levels, using only 20.9% of the time and 45.5% of the GPU memory compared with previous methods, while achieving SOTA performance on six tasks across nine datasets.

Related

• An AI4AI Framework for Visual Token Pruning
• HAP: Head-Adaptive Visual Token Pruning via Cross-Modal Alignment
• ERA: Entropy-Guided Visual Token Pruning with Rectified Attention for Efficient MLLMs
• Task-Related Token Compression in Multimodal Large Language Models from an Explainability Perspective
• LearnPruner: Rethinking Attention-based Token Pruning in Vision Language Models

Source: arXiv cs.CV | 2026-08-31