AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

MVPruner: Dynamic Token Pruning for Accelerating Multi-view Vision-Language Models in Autonomous Driving

Nan Yang, Zhanwen Liu, Linfeng Zhang, Shangyu Xie, Yang Wang, Wenzhuo Zhou, Xiangmo Zhao

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 MVPruner instead of Prune2Drive at 90% pruningDriveLMM-o1 · DriveLMM-o1Overall Reasoning: 70.31 -> 72.91Improved
use MVPruner instead of Prune2Drive at 90% pruningDriveMM · MAPLMAverage: 64.81 -> 68.30Improved
use MVPruner instead of Prune2Drive at 90% pruningDriveMM · DriveLMAverage: 57.4 -> 58.2Improved
use MVPruner instead of Prune2Drive at 90% pruningDriveMM · STSnuAverage: 43.17 -> 43.00No clear effect
use MVPruner at 75% pruningDriveMM · STSnuAverage: 44.21 -> 44.25Improved
use MVPruner instead of Prune2Drive at 90% pruningDriveMM · MAPLMFRM accuracy: 48.33 -> 53.27Improved
use MVPruner instead of Prune2DriveDriveMM · DriveLM efficiencydecoding time (ms): 59.9 -> 51.8Improved
use MVPruner instead of Prune2DriveDriveMM · DriveLM efficiencyprefilling time (ms): 307.5 -> 395.6Hurt
add second-stage pruning with diversity measure kept (Stage 1 alone applied to shallow layers) (ablation)MVPruner · DriveLM and MAPLM (90% pruning)accuracyImproved
remove future contribution metric from CCTS (ablation)MVPruner · DriveLM and MAPLM (90% pruning)accuracyHurt
use minimum pairwise distance instead of other diversity metrics in DRA (ablation)MVPruner · DriveLM and MAPLM (90% pruning)accuracyNo clear effect
place stage 1 at deeper layers (ablation)MVPruner · DriveLM and MAPLM (90% pruning)accuracyHurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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