AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

ModalTune: Fine-Tuning Slide-Level Foundation Models with Multi-Modal Information for Multi-task Learning in Digital Pathology

Vishwesh Ramanathan, Tony Xu, Pushpak Pati, Faruk Ahmed, Maged Goubran, Anne L. Martel

ICCV 2025 · Vision-language and multimodal › Cross-modal learning · approach: Medical multimodal fusion · 3 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 ViT-Adapter-style Modal Adapter instead of LoRA (ablation)ModalTune · TCGA subtypebalanced accuracy: 0.940 -> 0.954Improved
add ViT-Adapter-style Modal Adapter instead of LoRA (ablation)ModalTune · TCGA survivalC-index: 0.744 -> 0.750Improved
add transcriptomics tokens to the Modal Adapter (ablation)ModalTune · TCGA subtypebalanced accuracy: 0.920 -> 0.954Improved
add transcriptomics tokens to the Modal Adapter (ablation)ModalTune · TCGA survivalC-index: 0.707 -> 0.750Improved
add text-embedding multi-task training (ablation)ModalTune · TCGA survivalC-index: 0.730 -> 0.750Improved
add text-embedding multi-task training (ablation)ModalTune · TCGA subtypebalanced accuracy: 0.948 -> 0.954Improved
add text-embedding multi-task training (ablation)ABMIL (cat) · TCGA subtypebalanced accuracy: 1.6% dropHurt
add task-specific prompts to the general prompt (ablation)ModalTune · TCGA subtypebalanced accuracy: 0.935 -> 0.954Improved
add task-specific prompts to the general prompt (ablation)ModalTune · TCGA survivalC-index: 0.739 -> 0.750Improved
use CONCH text encoder instead of Llama-3-8B-UltraMedical (ablation)ModalTune · TCGA subtypebalanced accuracy: 0.921 -> 0.954Improved
add frozen random projector (ablation)ModalTune · TCGA survivalC-index: 0.730 -> 0.750Improved
make the projector trainable instead of frozen (ablation)ModalTune · TCGA subtypebalanced accuracy: 0.954 -> 0.652Hurt

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