AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

WSI-LLaVA: A Multimodal Large Language Model for Whole Slide Image

Yuci Liang, Xinheng Lyu, Wenting Chen, Meidan Ding, Jipeng Zhang, Xiangjian He, Song Wu, Xiaohan Xing, Sen Yang, Xiyue Wang, Linlin Shen

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Benchmarks and analysis of MLLMs · 33 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 WSI encoder (ablation)WSI-LLaVA · WSI-Bench testaverage score: 0.632 -> 0.680Improved
add WSI encoder (ablation)WSI-LLaVA · WSI-Bench morphological analysis openWSI-Relevance: 0.447 to 0.610Improved
add WSI encoder (ablation)WSI-LLaVA · WSI-Bench treatment planning openWSI-Relevance: 0.816 to 0.845Improved
add stage 1 WSI-text alignment on top of stages 2 and 3 (ablation)WSI-LLaVA · WSI-Bench testaverage score: 0.654 -> 0.680Improved
add stage 2 feature space alignment after stage 1 removed too (stage 3 only to full) (ablation)WSI-LLaVA · WSI-Bench testaverage score: 0.643 -> 0.654Improved
add all three training stages versus task tuning only (ablation)WSI-LLaVA · WSI-Bench testaverage score: 0.643 -> 0.680Improved
train/use WSI-LLaVA with full-WSI input instead of thumbnail MLLMsWSI-LLaVA vs Quilt-LLaVA · WSI-Bench testaverage score: 0.721 -> 0.754Improved
compare WSI-LLaVA against general-purpose GPT-4oWSI-LLaVA · WSI-Bench testaverage score: 0.507 -> 0.754Improved
WSI-LLaVA versus second-best on zero-shot histological gradingWSI-LLaVA · SlideBench-VQA BCNBhistological grading: 46.3% vs. second-best 30.0%Improved
WSI-LLaVA versus second-best on zero-shot molecular subtypingWSI-LLaVA · SlideBench-VQA BCNBmolecular subtyping: 29.2% vs. second-best 18.1%Improved
WSI-LLaVA versus Quilt-LLaVA on molecular subtypingWSI-LLaVA · WSI-Bench diagnosisWSI-Precision: 0.630 vs 0.675Hurt
use WSI metrics rather than NLU metrics for report evaluationHist-Gen vs WSI-LLaVA · WSI-Bench report caseWSI-Precision: BLEU-1 0.531 vs. 0.521; WSI-Precision 0.233 vs 0.8Mixed

Later papers that compared with this method

2 later accepted papers report WSI-LLaVA as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

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