AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

LLaVA-SP: Enhancing Visual Representation with Visual Spatial Tokens for MLLMs

Haoran Lou, Chunxiao Fan, Ziyan Liu, Yuexin Wu, Xinliang Wang

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants · 4 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 SP spatial tokens (Pooling variant) to LLaVA-1.5 trained with LoRALLaVA-1.5 Vicuna-7B · VizWizaccuracy: 45.7 -> 51.6Improved
add six visual spatial tokens (SFE, Cropping) (ablation)LLaVA-1.5 Vicuna-7B · 11-benchmark average (AvgN)AvgN: 63.4 -> 64.6Improved
add six visual spatial tokens (SFE, Pooling) (ablation)LLaVA-1.5 Vicuna-7B · 11-benchmark average (AvgN)AvgN: 63.4 -> 64.9Improved
use convolution instead of transformer blocks in SFE (Cropping) (ablation)SFE Cropping · 11-benchmark average (AvgN)AvgN: 64.5 -> 64.6Improved
use convolution instead of transformer blocks in SFE (Pooling) (ablation)SFE Pooling · 11-benchmark average (AvgN)AvgN: 64.4 -> 64.9Improved
use cropping tokens instead of sliding-window tokens (ablation)SFE · Spatial understanding subsetAvgN: 44.7 -> 46.5Improved
increase spatial tokens from 1 to 6 (ablation)SFE Cropping · 11-benchmark average (AvgN)AvgN: 63.4 -> 64.6Improved
increase spatial tokens from 6 to 12 (ablation)SFE Cropping · 11-benchmark average (AvgN)AvgN: 64.6 -> 64.6No clear effect
add DFI cross-attention (5x5 Zs-big) to SFE Cropping (ablation)LLaVA-SP-Cropping · 11-benchmark average (AvgN)AvgN: 64.6 -> 64.8Improved
add DFI cross-attention (5x5 Zs-big) to SFE Pooling (ablation)LLaVA-SP-Pooling · 11-benchmark average (AvgN)AvgN: 64.9 -> 65.1Mixed
try multi-layer CNN blocks or small convolution kernels in SFE (ablation)SFE · training stabilitytrainingHurt
add SP tokens (Cropping) for visual groundingLLaVA-1.5 · RefCOCO valaccuracy: 54.7 -> 60.3Improved

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