AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Structured Redundancy Modeling for Efficient Visual Token Pruning in High-Resolution MLLMs

Jouwon Song, Woohyeong Kim, Kyeongbo Kong

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Training-free token reduction for MLLMs · 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
add directional masking to SG Top-K selection (ablation)SFPruner · LLaVA-NeXT-7B ablation, 5x64 tokens (320)relative performance (%): 96.9 -> 97.9Improved
add directional masking to SG Top-K selection (ablation)SFPruner · LLaVA-NeXT-7B ablation, 5x64 tokens (320)selection time (ms): 0.7 -> 0.8Hurt
use directional masking instead of sequential selection (ablation)SFPruner · LLaVA-NeXT-7B ablation, 5x64 tokens (320)selection time (ms): 18.6 -> 0.8Improved
use directional masking instead of sequential selection (ablation)SFPruner · LLaVA-NeXT-7B ablation, 5x64 tokens (320)relative performance (%): 98.0 -> 97.9No clear effect
add SG-RLS to Top-K selection (replace SG) (ablation)SFPruner · LLaVA-NeXT-7B ablation, 5x64 tokens (320)relative performance (%): 96.9 -> 97.6Improved
add SG-RLS to directional masking (ablation)SFPruner · LLaVA-NeXT-7B ablation, 5x64 tokens (320)relative performance (%): 97.9 -> 98.3Improved
add SG-RLS to directional masking (ablation)SFPruner · LLaVA-NeXT-7B ablation, 5x64 tokens (320)selection time (ms): 0.8 -> 2.3Hurt
use SFPruner instead of CDPrunerSFPruner · Qwen2.5-VL-7B retain 40% tokensselection time (ms): 112.4 -> 2.5Improved
use SFPruner instead of CDPrunerSFPruner · Qwen2.5-VL MME stress test (9216 tokens, retain 30%)pruning time (ms): 576 -> 28Improved
apply SFPruner token pruning (30% retention)SFPruner · Qwen2.5-VL MME stress test (9216 tokens, retain 30%)prefill latency (ms): 1813 -> 1114Improved
apply SFPruner token pruning (30% retention)SFPruner · Qwen2.5-VL MME stress test (9216 tokens, retain 30%)total inference time (s): 6101 -> 3601Improved
use SFPruner instead of learned VisionSelectorSFPruner · Qwen2.5-VL-7B retain 20% tokensrelative performance (%): 95.7 -> 96.5Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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