AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

NaVLM-PVC: Progressive Visual Compression for Efficient Native-Resolution Encoding in MLLMs

Shichu Sun, Yichen Zhang, Haolin Song, Zonghao Guo, Chi Chen, Yidan Zhang, Yuan Yao, Zhiyuan Liu, Maosong Sun

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Attention-based visual token pruning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use global native-resolution encoding instead of slice-based encoding (ablation)SigLIP2 + Qwen2-7B MLLM · general benchmarks avgaverage accuracy: 67.5% -> 69.6%Improved
use global native-resolution encoding instead of slice-based encoding (ablation)SigLIP2 + Qwen2-7B MLLM · ShapeGrid spatial perceptionaverage accuracy: 62.6% -> 73.6%Improved
replace avg-pooling WTC with 2 stage compression (256 tokens) vs 1024 tokens baseline (ablation)ViT-UHD (SigLIP2) · Module ablationTTFT / OCR&Chart: TTFT 233 -> 82 ms; avg 62.1 -> 59.2Mixed
use content-adaptive pooling instead of plain average pooling in WTC (ablation)ViT-UHD (SigLIP2) · Module ablationaverage accuracy: 59.2 -> 60.7Improved
use pixel-unshuffle as WTC at early layers instead of pooling (ablation)ViT-UHD (SigLIP2) · Module ablationaverage accuracy: 60.7 -> 57.5Hurt
add refined patch embedding (patch 16 to 8) with 3 WTC (ablation)ViT-UHD (SigLIP2) · Module ablationaverage accuracy: 62.1 -> 63.0Improved
add more than 3 WTC layers or move their positions (ablation)LLaVA-UHD v3 · TTFT latencyTTFTNo clear effect
integrate PVC (RPE + WTC) into SigLIP2 backbone (ablation)SigLIP2 · PVC transferaverage accuracy: 61.94 -> 64.71Improved
integrate PVC into MoonViT backbone (ablation)MoonViT · PVC transferaverage accuracy: 66.17 -> 66.33No clear effect
apply same pre-alignment to MoonViT and ViT-UHD (PVC) (ablation)ViT-UHD · 8-benchmark average, max tokens 776 vs 380average accuracy: 64.11 -> 64.26No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy