transformer / foundation model

The paper identifies how medial frontal cortex integrates semantic content and sensory evidence for decision-making using human neurosurgical data and a vision-language model.

RadMultiBench evaluates pre-trained vision-language models on medical imaging tasks, showing performance gaps in report generation and classification across thoracic and breast imaging.

The paper compares transcriptomic profiles across kidney organoid, assembloid, and transplantation models to evaluate their maturity and suitability for research.

The paper shows that neural populations adjust their optimal temporal resolution during stimulus processing, becoming faster and more nonlinear when decoding information is most accurate.

This paper develops a 2D U-NET model to segment and quantify atherosclerotic lesions in mouse aortic sections, achieving high accuracy and significantly reducing measurement time.

VIGAT uses a vision-language model to quantify gene activity from single-cell ATAC-seq by analyzing accessibility patterns and genomic annotations across marker genes.

This paper identifies two failure modes in extended reasoning models within bioinformatics tasks and introduces DTDR to improve accuracy and reduce errors.

PhenoMe provides an open-source framework for analyzing microscopy images using foundation model-derived embeddings and experimental metadata to enable phenotyping and data exploration.

MaxCAT-Net improves skin lesion classification and segmentation with hybrid architecture and calibration, showing performance on benchmark datasets but with preliminary results.

This paper introduces a deep learning method that integrates genomic and trait data to improve species delimitation in complex speciation scenarios.

This study quantifies cancer stemness in prostate cancer across progression stages, showing its association with therapy resistance, poor survival, and lineage plasticity.

BioCluster uses foundation models to cluster and zero-shot label ecological images, achieving 84% and 83% species-level accuracy on moth and bird datasets.

SVC predicts subcellular transcript localization and co-organization using a Vision Transformer model, revealing disease patterns and spatial biology not accessible at the cell level.

SQUALL integrates histology and spatial transcriptomics to enable virtual spatial gene expression profiling and clinical prediction across cancers and platforms.

tRNAZAP is a deep learning toolkit for analyzing nanopore tRNA sequencing data, achieving high accuracy in segmentation and classification tasks.

Linear NeuroPaint evaluates a linear model for cross-session neural alignment and inpainting, finding it less accurate than nonlinear models but computationally efficient.

VesiScope detects and analyzes GUVs from phase contrast images using a YOLOv11 model, enabling accurate and efficient label-free quantification.

This paper develops a deep learning model to identify drug modes of action from high-throughput images of chemically and genetically perturbed Corynebacteria.

SAP provides a transferable, training-free framework for automated segmentation of plant time-series imagery across species and conditions.

This study evaluates two time-series foundation models for cognitive workload classification from eye-tracking data, showing better generalization to unseen individuals than task-specific models.

The paper introduces RSAUNet, a deep learning model for prostate cancer segmentation in MRI scans, achieving high Dice and Jaccard scores.

This paper compares segmentation methods for plant phenotyping, finding that pretrained vision transformers outperform color thresholding and supervised U-Nets in generalization across species and conditions.

The paper evaluates text-guided segmentation using SAM3 for weakly supervised leaf symptom segmentation, finding text guidance matches or outperforms conventional methods but is prompt-sensitive, with hybrid approaches improving robustness.

AdaGeneBudget reduces gene-token usage in single-cell foundation models while preserving annotation utility and fine-grained cell identities.

This study develops a sample-specific transformer model to improve genotype imputation from sparse single-cell RNA sequencing data, enhancing the recovery of genetic cell types.

QuantEM provides a vision transformer-based platform for EM data segmentation and analysis, achieving accurate zero-shot segmentation with less data than existing models.

Deep-Pose-Tracker automates C. elegans behavior analysis by detecting poses and quantifying locomotion and posture features from videos.

RVQ-Alpha adapts large language models for single-cell analysis by integrating hierarchical tokenization and reinforcement learning, achieving performance comparable to task-specific models on a multi-task benchmark.

This paper develops a computer vision approach using YOLO models to automate fecundity assays in C. elegans, achieving high accuracy and reducing human bias compared to manual counting.

The paper develops MAESTRO, a self-supervised deep learning framework to transform immune profiles into embeddings capturing immune health, enabling clinical prediction and precision immunotherapy.