Research Applications

This use case focuses on accurately classifying Isocitrate Dehydrogenase (IDH) mutation status using computer-aided pathology to improve diagnosis and treatment of adult-type diffuse gliomas.

Using data-based virtual sensors and causal physics features to estimate solenoid position accurately under concept shifts when high-fidelity models are unavailable or costly.

This study developed an interpretable AI framework integrating cholangiographic imaging and clinical data using TabPFN to support postoperative risk stratification of recurrence and all-cause mortality after ERCP-guided stone extraction in choledocholithiasis…

This study systematically evaluates machine learning models, including TabPFN, for predicting post-wildfire debris flows using basin-scale observations across the western United States. It focuses on model performance, feature importance, and the effect of sy…

The study presents a machine-learning pipeline based on a tabular foundation model to identify cognitive impairment risk across diverse countries using indicators of reduced survey response quality from older adults answering psychosocial questionnaires. The…

This study empirically evaluates and compares Real-TabPFN against representative AutoML pipelines for cardiovascular disease prediction, focusing on the extension of TabPFN through continued pre-training on real-world datasets to improve synthetic-to-real gen…

Evaluating the effectiveness of general-purpose Tabular Foundation Models such as TabPFN against specialized models in predicting cellular responses to genetic and chemical perturbations across various biological datasets and scales.

Optimizing catalyst formulation and operating parameters for the water–gas shift reaction to improve CO conversion, using an interpretable TabPFN-ensemble framework.

Applying patient-specific computational tools, including machine learning and mechanistic models, to improve personalization of treatment strategies for acute respiratory failure across various patient populations and clinical settings.

An AI-driven framework combining TabPFN and AutoGluon predicts and optimizes wheat yield and nitrogen uptake using global field trial data. The approach identifies management strategies that increase average yield by 19.6% while reducing nitrogen application…

This study uses features extracted from Transcranial Doppler (TCD) to estimate chronological brain vascular age and assess accelerated aging in subjects with various brain diseases, utilizing machine learning regression models trained on healthy subjects.

This research proposes PromptDx, a novel diagnosis-by-reference framework improving upon In-Context Learning approaches like TabPFN for Alzheimer’s diagnosis using multimodal data.

The study develops and evaluates machine learning-based counterfactual outcome models trained on IFX-treated patients to predict treatment outcomes for ADA-treated pediatric Crohn’s disease patients, enabling virtual control arms in single-arm inflammatory bo…

TACO is a framework combining causal inference with the TabPFN foundation model to detect Long COVID presymptomatically. It uses Differential Causal Effect analysis to identify causally relevant genes and employs TabPFN for consistent performance without hype…

Using tabular foundation models (TFMs) to perform in-context learning for molecular property prediction in drug discovery, catalysis, and chemical engineering, especially effective in low- to medium-data regimes without task-specific training.

This work introduces TimeSeriesExamAgent, a framework to scale creation of comprehensive time series reasoning benchmarks using synthetic and real-world data from healthcare, finance, and weather domains. It evaluates large language models (LLMs) across multi…

A large-scale benchmark study comprising 3030 tabular datasets across diverse industries to evaluate and compare state-of-the-art models including gradient boosting decision trees, neural networks, and foundation models, providing actionable guidance on which…

A zero-shot prediction pipeline combining BulkFormer and TabPFN v2.5 predicts delayed kidney function directly from raw RNA-Seq data without model retraining. Tested on human kidney transplant cohorts, the approach achieves strong predictive performance and d…

An interpretable machine learning framework using TabPFN predicts reaction rate constants for antibiotic degradation catalyzed by biochar. The model outperforms other ML techniques with high accuracy and interpretability, revealing key catalyst properties and…

This study develops a machine learning ecosystem to improve cardiovascular disease diagnosis through early warning systems using self-reported features and specialized diagnostic models integrating clinical and non-clinical data. It uses advanced ML methods i…

Development and validation of a multimodal AI model using transformer-based TabPFN to estimate appendicular lean mass (ALM) and detect low muscle mass from frontal and lateral chest radiographs and electronic health record (EHR) data, demonstrating robustness…

Predicting recovery outcomes after spinal cord injury using data-driven modeling, with a focus on motor status and walking function prediction based on early clinical assessments from the Sygen trial dataset.

The study developed a transcriptome-based approach combining bulk RNA-seq data from kidney cell lines, primary cells, or tissues with single-cell RNA-seq reference datasets to accurately match and validate kidney cell types. TabPFN, alongside Spearman correla…

A method for survival analysis in medical applications using a prior-fitted in-context learning model pretrained on synthetic data to enable individualized survival prediction without task-specific training.

This study presents a fully automated methodology for early prediction studies in clinical settings, leveraging information extracted from unstructured discharge reports. The proposed pipeline supports cohort selection, dataset generation, and outcome labelin…

This paper presents an interpretable machine-learning framework for predicting the splitting strength (ST) of asphalt concrete and supporting data-driven mixture design. A database consisting of 296 samples was established, and 14 input variables related to a…

This study uses the TabPFN model to predict the International Roughness Index (IRI) of asphalt pavements by integrating traffic load, pavement structure parameters, and climate variables, outperforming traditional regression models and other machine learning…

This use case addresses water inrush risks in coal mining by integrating 2D discrete-element simulation (PFC2D) with TabPFN for small sample prediction and SHAP for explainability, enabling accurate and interpretable prediction of water-conducting fracture zo…

A TabPFN-based prediction and optimization framework designs sustainable cemented foam backfill by balancing mechanical strength, cost, and carbon emissions. Combined with SHAP for interpretability and NSGA-II for multi-objective optimization, the approach id…

This dataset contains high-resolution industrial carbon emissions data for China from 2000 to 2020, generated using a machine learning framework based on land-use probability and geospatial datasets. The data is provided at a 30-meter resolution and covers in…

This study evaluates the performance of the TabPFN model compared to classical ensemble methods for EEG seizure segment classification, showing TabPFN’s superior accuracy in cross-subject generalization.

The use case involves predicting mechanical properties of steel during hot rolling processes such as Thin Slab Direct Rolling (TSDR). It uses a multitask learning framework that enhances TabPFN with multitask-aware prior adaptations to capture correlations am…

A TabPFN-based framework predicts the shear strength of fault gouge under challenging small-sample and missing-data conditions. By combining incomplete fault gouge data with auxiliary geological samples, TabPFN improves predictive accuracy by 19.3% over the M…

TabPFN is applied to early-stage drug discovery where data scarcity and out-of-distribution shifts limit the reliability of predictive models. It offers robust and data-efficient tabular modeling, outperforming traditional methods like XGBoost in regression t…

Application of TabPFN and its tabpfn-extensions library to geotechnical tasks including soil-type classification and iterative imputation of mechanical parameters, with visualization and uncertainty quantification aiding interpretability in data-scarce scenar…

A condition-adaptive cloud-edge collaborative framework named CAPE is proposed for photovoltaic power forecasting, combining site-specific expert models, lightweight edge-side models, and cloud-based large retrieval models to enhance forecasting accuracy, rob…

A systematic multi-cohort EHR benchmark evaluates TabPFN-based in-context learning against classical and deep learning models under realistic clinical challenges such as limited data, high dimensionality, and severe class imbalance. The study introduces AWARE…

FinTRACE is a retrieval-first architecture that converts raw transaction data into reusable feature representations for improved user modeling in financial organizations. It uses rule-based detectors and stores signals in a behavioral knowledge base to enhanc…

Development of a delay-aware digital-twin MPC benchmarking framework for anaerobic digesters using various surrogate ML models including TabPFN to provide fast one-step predictions under irregular measurements, enabling improved methane production and robust…

A knowledge-constrained data augmentation framework combines domain-informed synthetic data generation with TabPFN to improve concrete property prediction across 15 datasets. By detecting and removing unrealistic synthetic samples, the approach improves predi…

This use case involves using TabPFN as a surrogate model in active learning for materials discovery. TabPFN replaces conventional Gaussian Process and Random Forest models to provide better uncertainty calibration and regression performance, enabling more dat…

Development of a machine learning framework for postoperative breast cancer related lymphedema (BCRL) risk prediction using routinely collected clinical data from breast cancer patients. Various classifiers, including transformer-based TabPFN, were evaluated…

This study proposes a symmetric leaf-vein (SLV) texture to improve thermohydrodynamic lubrication performance of journal bearings. It uses a surrogate-assisted multi-objective optimization combining a TabPFN surrogate model with NSGA-II to optimize texture pa…

A leakage-aware framework for early detection of rug pull attacks in blockchain ecosystems using on-chain behavioral metrics and temporally aligned OSINT signals, employing TabPFN for modeling multimodal tabular data under temporal constraints.

This study uses AlphaFold3 and machine learning to predict 3D structures and coupling specificity of human GPCR-G protein signaling complexes, providing a comprehensive computational atlas to understand signaling mechanisms and enable precision therapeutics.

This use case involves predicting the maximum loading of alkane isomers in zeolites for hydroisomerization processes using a cascade machine learning model including a TabPFN pretrained model. The model improves prediction accuracy of adsorption capacity thro…

This research presents an efficient method for credal prediction that produces intervals representing plausible probabilities for class labels, enabling uncertainty representation in machine learning. The method is computationally efficient, overcoming challe…

Development of an accurate, low-cost, and clinically feasible early screening model to identify women at high risk of recurrent spontaneous abortion (RSA) using routinely available clinical biomarkers, leveraging a Transformer-based tabular model (TabPFN) alo…

Using TRAMs derived from delayed-contrast MRI and radiomic features to distinguish malignant breast lesions from benign ones, improving diagnostic precision and potentially reducing unnecessary biopsies in breast cancer diagnosis.

This work introduces TREDBench, a collection of 83 real-world tabular regression datasets with engineering labels, to study domain structure using TabPFN 2.5’s dataset-level embedding. It identifies a domain gap between synthetic and engineering data and prop…

Using a multimodal multi-instance learning framework to predict peak oxygen consumption (peak VO₂) from transthoracic echocardiography studies and electronic health records, improving accuracy and generalization for heart failure patient assessment.

This study develops a machine learning framework using TabPFN to assess slope stability based on geometric features and shear strength parameters. It integrates field investigations with numerical simulations to create a 2D slope database. TabPFN outperforms…

This use case involves the development and application of transformer-based probabilistic deep learning models to predict the success of hematopoietic stem cell (HSC) mobilization in healthy donors using baseline and post-mobilization laboratory data. The mod…

This use case focuses on predicting the fracture energy (Gc) and Mode I fracture toughness (KIC) of concrete structures using 25 machine learning algorithms including classical models and modern transformer-inspired neural networks like TabPFN. TabPFN outperf…

This study evaluates the effect of class-imbalance correction techniques on discrimination and calibration of clinical risk prediction models across diverse medical datasets.

This use case addresses the classification of adolescent alcohol drinking behavior using low-cost, accessible clinical measures rather than expensive neuroimaging data. The proposed framework, FocalTab, integrates TabPFN with focal loss to improve accuracy an…

TabPFN is evaluated against gradient-boosted trees across three solar energy meteorology tasks, including irradiance bias correction, irradiance decomposition, and solar power forecasting. TabPFN consistently outperforms gradient boosting, with particularly s…

A large-scale benchmark evaluates TabPFN and microbiome-specific foundation models for disease prediction across 83 cohorts and 20 diseases. TabPFN provides strong out-of-the-box performance and competitive robustness across cohorts, but does not consistently…

Using TabPFN, a foundation model pretrained on synthetic classification tasks, adapted via ensemble class balancing and domain-specific finetuning, to improve species distribution models (SDM) by surpassing traditional methods in discrimination and calibratio…

This study utilized UAV multispectral imagery and in situ soil salinity samples to create a stacking ensemble model incorporating TabPFN, SVM, and Ridge regression with XGBoost as meta-learner to accurately monitor soil salinity in coastal wetlands, demonstra…

An explainable TabPFN model predicts the geographical origin of kelp in China using stable isotope measurements. TabPFN achieves 100% classification accuracy and an AUC of 1.0 in just 2.5 seconds, providing a 5,782-fold speedup over tuned traditional machine…

MultiModalPFN (MMPFN) extends TabPFN to integrate heterogeneous data modalities such as images and text alongside tabular data, improving performance in healthcare and general-purpose multimodal datasets by providing a unified, scalable framework for heteroge…

ChemPFN adapts the TabPFN approach to molecular bioactivity prediction, learning from over 10 million ChEMBL measurements to make data-efficient predictions for small molecular assays. Using labeled molecules as context, it provides well-calibrated uncertaint…

An explainable TabPFN model predicts CO₂ frosting temperature in natural gas mixtures under diverse operating conditions. TabPFN outperforms CatBoost and optimized LSSVM with an R² of 0.9992 and MAPE of 0.13%, while SHAP analysis identifies CO₂ concentration,…

This approach combines the Conditional Randomization Test (CRT) with TabPFN to provide valid hypothesis testing and p-values for assessing feature relevance in tabular data, enabling finite-sample valid p-values without model retraining or parametric assumpti…

This study proposes a surrogate modeling framework combining Autoencoder and TabPFN to predict the internal temperature field of rotary kilns, enabling efficient reconstruction and real-time monitoring with high accuracy, overcoming data scarcity and signific…

A TabPFN-based clinical model distinguishes psychotic major depression (PMD) from non-psychotic major depression (NPMD) using electronic medical record data. Using eight LASSO-selected features, TabPFN outperforms traditional machine learning models with an A…

Development of ¹H NMR-based metabolomics-AI platforms integrating serum metabolomic profiles and clinical parameters to enable rapid, accurate, and non-invasive early detection of pancreatic ductal adenocarcinoma (PDAC).

Using TabPFN to approximate conditional expectations for Shapley value computation in explainable AI, achieving better or comparable performance to state-of-the-art methods with lower runtime.

Machine learning models were trained to predict adverse events requiring hospitalization and forecast AE onset timing for optimized ward management in myeloma patients undergoing stem-cell mobilization and collection, aiming to enable a risk-stratified outpat…

Proposes a zero-shot generation framework for parametric engineering design based on TabPFN, enabling conditional design generation with limited reference samples and no task-specific training. The method generates design parameters sequentially based on targ…

ExplainerPFN is a tabular foundation model built on TabPFN, pretrained on synthetic structural causal datasets supervised with exact or near-exact Shapley values, designed to predict feature attributions for unseen tabular datasets without model access, gradi…

This work investigates the use of TabPFN, a learning-free model, for malware detection in cybersecurity under limited data conditions. It compares TabPFN’s performance against established ensemble models such as Random Forest, LightGBM, and XGBoost across mul…

This use case presents a predictive method combining an Autoencoder, Random Forest feature selection, and TabPFN to accurately diagnose prostate and breast cancer using metabolomics data acquired from advanced mass spectrometry techniques. The model achieved…

The study presents a unified in-context learning foundation model framework that couples a pretrained TabPFN transformer with graph neural network embeddings and novel MagpieEX descriptors for materials property prediction. This approach achieves competitive…

This work proposes a pre-trained model based framework for predicting cellular toxicity of silica nanoparticles. It addresses data leakage by removing evaluation stage features and uses the embedding layer from TabPFN to convert categorical values into dense…

This study uses AI models including TabPFN to predict and optimize the compressive strength of fly ash-based geopolymer concrete, achieving high accuracy and stability in predictions and combining metaheuristic algorithms for mix design optimization.

This study investigates the use of neural networks and transfer learning models, including TabPFN, to predict mycotoxin contamination in Irish oat crops. Using a dataset with environmental, agronomic, and geographical predictors, various models were evaluated…

Using a descriptor-first strategy combined with a pretrained foundation model for tabular data (TabPFN) for single-task regression in predicting ADMET properties relevant to small-molecule drug design, achieving improved accuracy and simplified deployment com…

Investigating whether graph node classification can be reformulated as a tabular learning problem using TabPFN-GN, which converts graph data into tabular features to enable direct node classification, achieving competitive or better performance than GNNs.

Study on using early arterial lactate trajectory patterns to predict mortality risk in ICU-admitted heart failure patients, identifying distinct lactate classes that correlate with different mortality outcomes and providing enhanced prognostic insights for ea…

This study evaluates the use of TabPFN for modeling volatile organic compound (VOC) profiles obtained via mass spectrometry methods like gas chromatography–mass spectrometry and proton transfer reaction mass spectrometry. TabPFN is applied to high dimensional…

This use case describes using AI peer review with large language models (LLMs) and tabular models to accelerate research and improve performance in glioblastoma clinical trial patient matching. The study replaced earlier deep learning notebooks with a TabPFN-…

Development and validation of an AI predictive model using clinical data to forecast post-stroke recurrence, disability, and mortality at multiple time points, facilitating targeted interventions for high-risk stroke patients.

This study benchmarks nine open-source AutoML frameworks and the transformer-based model TabPFN across 25 financial predictive analytics datasets, evaluating their performance in various financial tasks. It finds that TabPFN excels in multiclass tasks within…

This use case applies TabPFNv2 in analog circuit optimization by leveraging pre-trained tabular models encoding device physics primitives, enabling reliable optimization without per-circuit engineering. The approach, called Circuit Prior Network (CPN), combin…

This use case involves applying various machine learning techniques to predict the compressive strength of geopolymer concrete, using a dataset of 204 instances including experimental data. Features such as chemical composition and curing temperature are used…

This study evaluates how measurement errors in respiratory rate and heart rate affect the accuracy of predicting the success or failure of high flow nasal cannula (HFNC) therapy in patients with acute hypoxemic respiratory failure. It compares the TabPFN mach…

A TabPFN-based model predicts hepatic encephalopathy following TIPS by integrating CT radiomics, deep transfer learning features, and clinical data. The model achieves strong externally validated performance with an AUC of 0.942, while SHAP analysis identifie…

This study developed a method to evaluate driver alertness on mountain roads by integrating human and environmental factors. Data such as heart rates, eye movements, and demographics were collected to classify alertness levels. TabPFN model was used alongside…

The study uses the Optuna-TabPFN model to accurately predict the unconfined compressive strength (UCS) of cemented paste backfill (CPB) with limited data, achieving high accuracy and employing SHAP analysis to interpret the model and identify key influencing…

Use of TabPFN as the surrogate model within a Meta-Black-Box Optimization framework (DB-SAEA) to enhance surrogate-assisted evolutionary algorithms for multi-objective optimization problems, enabling accurate prediction and uncertainty estimation, and improvi…

This study introduces a bias-aware and accurate predictive framework integrating TabPFN with Monte Carlo Bias Estimation (MCBE) to model the elastic modulus of recycled aggregate concrete (RAC), enabling reliable, bias-corrected, and interpretable predictions…

This study developed a scenario-aware modeling framework for the trophic level index (TLI) of small lakes by integrating multi-source satellite imagery and machine learning algorithms including TabPFN Regression. The approach enabled seamless monitoring of eu…

A hybrid approach using TabPFN combined with the Coulomb and Proximity Potential Model (CPPM) to study α-particle preformation factors and their impact on α-decay half-lives, improving predictions and identifying nuclear structure correlations.

A deep-learning surrogate model based on TabPFN predicts free energy with high fidelity to assess synthetic feasibility of hypothetical metal-organic frameworks (hMOFs), enabling rapid flagging of prime MOF candidates for synthesis and accelerating free-energ…

This research applies machine learning models, including TabPFN, to predict crack mouth opening displacement (CMOD) in fiber-reinforced ultra-high-performance concrete (FR-UHPC). The study uses a comprehensive experimental database and compares several ML alg…

PSF-Net combines TabPFN’s meta-learned prior knowledge and SAINT’s dual attention mechanism with an uncertainty-aware gating block to predict downlink power density and classify radiation levels from 5G base stations. The model provides calibrated uncertainty…

Improving the accuracy of predicting intra-European genetic ancestry by selecting more informative genetic markers through machine learning methods, outperforming traditional marker panels in forensic genetics.

Benchmarking different methodologies including count-based models with TabPFN, pretrained transformers, and mixture-of-agents LLM pipelines for clinical prediction on structured electronic health records using the EHRSHOT dataset.

This study uses an automated supervised classification approach combining TabPFN with SHAP-based interpretability to accurately detect Thunderstorm Ground Enhancements (TGEs) from a dataset of 2024 events. The framework achieves high classification accuracy a…

TabMGP is a Martingale Posterior (MGP) method built on TabPFN designed for Bayesian inference on tabular data. It provides credible sets with near-nominal coverage and often outperforms handcrafted MGP constructions and standard Bayesian baselines. It enables…

This use case involves using TabPFN, a pre-trained transformer model for tabular data classification, to differentiate between benign and malignant renal tumors based on CT radiomics datasets. Compared with conventional machine learning algorithms, TabPFN sho…

This study develops machine learning models, including TabPFN, to predict substantial hypo- and hyperglycemia on dialysis days in patients with diabetes using continuous glucose monitoring data and baseline characteristics.

The study uses artificial intelligence frameworks, including the TabPFN model, to distinguish and classify corneal stromal and endothelial disorders from controls and to categorize corneas into seven groups for diagnosis with high accuracy.

OBJECTIVETo construct a classification prediction model for vancomycin blood concentration， and to optimize its precision dosing strategies.METHODSPatient records meeting inclusion criteria were extracted from the Medical Information Mart for Intensive Care…

Using machine learning models augmented with TabPFN data augmentation to predict capacity and initial Coulombic efficiency of biomass-derived hard carbon for sodium-ion batteries, identifying key synthesis parameters and improving screening efficiency.

The study proposes an approach using tabular foundation models for accurate spatiotemporal forecasting of small Earth data, which shows superior accuracy compared to traditional graph deep learning and TabPFN models in most scenarios.

This use case focuses on predicting ship fuel consumption using various modeling approaches, including tabular and time-series regression models. It benchmarks these models on a new dataset comprising operational and environmental data from three ships and in…

TabPFN-Wide extends prior-data fitted networks through continued pre-training to handle datasets with extremely high feature counts (over 30,000), maintaining interpretability and robustness to noise for biomedical omics data.

The paper proposes AttenMLP, which integrates attention-based feature replay with context retrieval and sliding buffer updates for streaming tabular learning. It aims to improve energy and memory efficiency in continual learning for tabular data streams with…

A novel foundation model-based recommendation framework using TabPFN to integrate multi-omics profiles with clinical factors for personalized neoadjuvant therapy recommendations in breast cancer, improving treatment response rates significantly.

Abstract Long COVID affects 10-40% of COVID-19 survivors, yet early detection remains challenging. We present TACO (TabPFN Augmented Causal Outcomes), a framework that uniquely combines causal inference with foundation models for presymptomatic Long COVID det…

TabImpute is a pre-trained transformer model built on TabPFN for universal zero-shot imputation of missing data in tabular datasets. It requires no fitting or hyperparameter tuning at inference, significantly speeding up imputation and improving accuracy, esp…

This study developed and clinically validated an interpretable Transformer-based predictive model, TabPFN-V2, to assess gastrointestinal bleeding risk in pediatric patients with Henoch-Schönlein purpura (IgA vasculitis). The model used comprehensive clinical…

ObjectiveHenoch-Schönlein purpura (HSP), clinically recognized as IgA vasculitis (IgAV), a prevalent systemic vasculitis in pediatric populations, frequently…

This study investigates an uncertainty-guided strategy for model selection using TabPFN models in biomolecule efficacy prediction, specifically on an siRNA knockdown efficacy task. The proposed OligoICP method selects an ensemble of TabPFN models with the low…

This study evaluated the predictive performance of various machine learning and deep learning models including TabPFN in estimating manual strength in men and women using anthropometric data collected from 382 participants. TabPFN and TabNet deep learning mod…

Development of a machine learning framework to rapidly and accurately predict Henry coefficients of linear and branched alkanes in one-dimensional zeolites, enabling efficient screening and design of shape-selective zeolite catalysts to support catalytic hydr…

Abstract Hematopoietic stem and progenitor cell (HSPC) transplantation offers a potentially curative therapy for aggressive hematologic malignancies and bone marrow failure syndromes. Successful transplantation depends on effective mobilization of donor CD34…

BackgroundNegative emotionality is a core dimension of infant temperament, characterized by heightened distress, reactivity, and difficulty with self-regulat…

Asmaa A. Mahdi Accurate prediction of stroke patient status is vital for early intervention and clinical decision-making. This study systematically evaluates…

This study develops a deep tabular learning framework to predict crash severity in electric vehicle collisions using Texas real-world crash data (2017-2023). It addresses class imbalance with SMOTEENN resampling and benchmarks TabPFN alongside MambaNet and Ma…

Evaluating the performance of Variational Bayesian Last Layers (VBLL) integrated with TabPFN for uncertainty estimation in medical tabular datasets, comparing with original TabPFN.

A hybrid framework called kriging prior regression (KpR) that enhances machine learning with spatial context using spatial lag features from ordinary kriging, applied to digital soil mapping. The approach uses the TabPFN model to improve predictions and uncer…

ObjectiveTo develop a self-reportable risk assessment tool for elderly type 2 diabetes mellitus (T2DM) patients, evaluating risks of diabetic nephropathy (DN…

Development of a multi-modal machine learning framework that integrates cell-free DNA structural features and genetic information for early prediction of gestational diabetes mellitus at 10-12 weeks gestation in Vietnamese women. The approach uses TabPFN clas…

A machine learning study predicts tuberculosis treatment outcomes using over two decades of data from Brazil’s national SINAN-TB database. While Random Forest achieves the best overall performance, TabPFN delivers competitive results despite data-volume const…

Tuberculosis (TB) remains a significant public health concern, particularly in low- and middle-income countries such as Brazil. Predicting treatment outcomes is essential to guide clinical decisions and strengthen public health strategies. Thus, this study ev…

Application of TabPFN to geotechnical site characterization problems including spatial variation prediction of undrained shear strength and imputation of missing mechanical parameters using zero-training few-shot learning with added context from BID.

This work evaluates the use of Large Language Models (LLMs) for data fitting tasks, highlighting their vulnerability to task-irrelevant variations such as changes in variable names. It compares LLMs to the TabPFN tabular foundation model, noting that while Ta…

This work introduces L2C-TabPFN, a method integrating longitudinal-to-cross-sectional transformation with a pre-trained TabPFN model to predict Alzheimer’s disease outcomes using multimodal clinical data from the TADPOLE dataset. It enables robust prediction…

A hybrid AI/PBPK framework predicts both molecular dissolution and systemic absorption of amorphous solid dispersion (ASD) drug formulations. TabPFN outperforms five other machine learning approaches for dissolution prediction with an R² of 0.905, while integ…

Benchmark study evaluating multiple neural network architectures including TabPFN for predictive soil modeling at field-scale digital soil mapping, showing TabPFN’s strong performance and robustness on small tabular datasets.

This study evaluates the effectiveness of solar-reflective cool paint in tropical urban environments by combining a transient physical model and data-driven machine learning models including TabPFN regressor to predict facade temperatures and heat fluxes.

This use case focuses on applying small-data machine learning models such as TabPFN for forecasting individual-level mental health trajectories using personalized digital phenotyping data. It addresses challenges of sparse, irregular, and noisy data in contin…

Development and validation of an interpretable transformer-based clinical prediction model (TabPFN-V2) for identifying IVIG resistance in pediatric Kawasaki disease patients, demonstrating superior predictive performance and interpretability to support person…

Classifying blazars from the Fermi 4LAC-DR3 catalogue to identify origins of blazars of uncertain type and studying their structure and redshift-luminosity evolution using ML models including TabPFN.

Application of TabPFN for sub-national crop yield forecasting in South Africa using Earth Observation and weather time series data, showing comparable accuracy to ML models but with faster tuning and less feature engineering required.

The study applies the TabPFN algorithm to combine delta-radiomics features derived from dynamic MRI with clinical and morphological features to improve the preoperative prediction of lymphovascular invasion in invasive breast cancer patients.

This study evaluates the performance of three advanced tabular deep learning models—MambaAttention, TabPFN, and TabTransformer—for classifying SAE automation levels in automated vehicle crashes. It uses structured crash data from Texas (2024) involving 7,300…

Method to detect empathy from video interactions using TabPFN v2 and TabICL under strong privacy constraints by leveraging summary statistics of temporal visual features. Achieves strong utility and improves generalization on a public human-robot interaction…

This use case involves the application of a novel framework combining tabular-tailored dataset distillation with pretrained models like TabPFN to enhance credit scoring technologies in the financial sector. The framework addresses class imbalance during datas…

The hydroisomerization of long-chain alkanes is a key catalytic process for producing high-quality fuels and lubricants, where zeolites play a central role due to their shape-selective properties. However, the vast diversity of zeolite frameworks and alkane i…

Development of non-invasive predictive models for diagnosing minimal change disease (MCD) using machine learning algorithms including TabPFN. The study compares TabPFN’s performance against other models and identifies key predictors using SHAP.

하천에서 조류의 과다 발생은 취수원 관리 및 정수 처리에 악영향을 줄 수 있어 지속적인 관리가 필요하다. 본 연구에서는 딥러닝 알고리즘 중 작은 규모의 테이블 데이터에서도 상대적으로 우수한 성능을 보이는 것으로 알려진 tabular prior data fitted networks (TabPFN)을 사용하여 조류 발생 지표 중 하나인 chlorophyll-a (chl-a) 농도를 예측하는 다중 분류 모형을 구축하였다. 모형의 구축을 위해 부여지점 수질자동측정망에…

Efforts to use open source practices to attract data scientists for collaborative antimalarial drug discovery, enhancing reproducibility, progress tracking, and inclusivity in drug discovery.

A novel framework using the TabPFN model to forecast blockwise grapevine diseases by leveraging climate variables from multi-sensor remote sensing imagery, enhancing accuracy and efficiency in disease prediction and enabling precise, targeted interventions in…

TabPFGen uses TabPFN as an energy-based generative model to effectively generate tabular data without additional training or hyperparameter tuning, showing strong performance in data augmentation, class-balancing, and imputation.

Chemical modulation of proteins enables a mechanistic understanding of biology and represents the foundation of most therapeutics. However, despite decades of research, 80% of the human proteome lacks functional ligands. Chemical proteomics has advanced fragm…

In the ever-evolving landscape of insurance services, the strategic cross-selling of health insurance to existing customers stands as a pivotal endeavor. This study delves into this domain, aiming to harness the potential of machine learning and deep learning…

The study develops a hybrid LightGBM-TabPFN model to predict dementia conversion in Parkinson’s disease patients using a small dataset with missing values. The model achieves high accuracy and interpretability with SHAP analysis, providing insights for clinic…

Intelligent fault detection and classification is a cornerstone of prognostic and health management of rotating machinery (RM) research. Correctly classifying and predicting RM faults not only increases productivity in industrial plants but also reduces maint…

A TabPFN-based clinical prediction framework estimates short-term outcomes following anterior cervical corpectomy, including prolonged hospital stay, non-home discharge, and major complications. The models achieve AUROCs of up to 0.816, while SHAP identifies…

A machine learning framework predicts shallow natural gas hazards in tunnels by first clustering historical cases into distinct disaster-causing patterns. Combining K-means clustering with TabPFN improves classification accuracy by 4% over an unclustered appr…

A forensic genetics study evaluates TabPFN against established ancestry classification methods for predicting biogeographical ancestry from genetic data. TabPFN consistently achieves the best accuracy, ROC AUC, and log loss for both continental and intraconti…

A comparative study evaluates nine machine learning models for predicting tilt in ultra-deep caissons using denoised construction data. TabPFN achieves the best performance with R² values above 0.99 and remains highly accurate even with only 10% of the traini…

A TabPFN-based surrogate modeling and optimization framework improves the energy efficiency of multistage centrifugal pumps by predicting pump head and entropy generation. Combined with SHAP and NSGA-II optimization, the approach reduces total entropy generat…

A comparative machine learning study evaluates TabPFN and 13 other models for predicting stillbirth from cardiotocography data. TabPFN achieves the best performance with 97.91% accuracy and an AUC of 98.88%, demonstrating strong potential for early identifica…

AutoEnergy is an automated, domain-aware feature engineering method for energy consumption forecasting, evaluated across 18 real-world energy datasets. It reduces forecasting errors by 19.5–84.7% compared with existing approaches while requiring less manual f…

A prediction and optimization framework combines TabPFN with NSGA-II to design sustainable cemented foam backfill under small-data conditions. The approach balances compressive strength, cost, and carbon emissions, while SHAP reveals key factors controlling s…

A comparative study evaluates machine learning and deep learning models for predicting manual strength from anthropometric measurements in 382 adults. TabPFN and TabNet achieve strong performance for torque strength by capturing nonlinear relationships, while…

A large retrospective study evaluates CorvisST biomechanical indices and AI models for diagnosing and differentiating seven corneal conditions across 1,500 eyes. TabPFN achieves the highest overall diagnostic accuracy at 88.7%, while key biomechanical indices…

A study investigates cryogenically assisted abrasive waterjet machining of Ti-6Al-4V alloy using liquid nitrogen to improve surface integrity and reduce thermal damage. A Bootstrap + TabPFN model predicts surface integrity with an R² of 0.955, complementing e…

An uncertainty-aware framework improves real-time detection of adverse geological conditions during tunnel boring machine excavation. An enhanced spatiotemporal TabPFN achieves R² values above 0.97 for multi-step operational forecasting, while Monte Carlo sim…

An interpretable machine learning study evaluates five models for predicting coal grindability across two large datasets. TabPFN achieves the best performance with R² values of 0.883 and 0.738, outperforming Random Forest, GBRT, SVM, and XGBoost, while SHAP,…

A clinical study evaluates TabPFN against eight traditional machine learning methods for predicting coronary stenosis in patients with cardiovascular-kidney-metabolic syndrome and coronary heart disease. TabPFN performs best under small-sample conditions, ach…

A TabPFN-based modeling framework evaluates how artificial accelerated aging can approximate the long-term effects of natural aging on biochar-treated, cadmium-contaminated soils. TabPFN provides strong predictions of soil active cadmium and pH, enabling regi…

An explainable TabPFN-based surrogate model predicts thermal performance in nano-enhanced phase change material battery thermal management systems. TabPFN outperforms BPNN, XGBoost, and CatBoost with up to a 92% reduction in prediction error, while SHAP analy…

A lightweight and interpretable AI framework combines lesion segmentation with TabPFN-based classification for skin cancer diagnosis. By fusing image-derived segmentation features with clinical data, the system achieves state-of-the-art performance on ISIC 20…

A machine learning study predicts the mineralogical and mechanical properties of three lunar meteorites using only their atomic composition. A prior-data fitted network achieves near-perfect classification of meteorites and minerals, while regression models a…

FinPFN is a Transformer-based meta-learning framework for dynamic stock return prediction that adapts forecasts to changing market regimes using recent feature-return relationships. Evaluated on Chinese A-shares and U.S. equities, FinPFN significantly outperf…

MuCB-tabpfn is a multimodal TabPFN-based framework that predicts human blood concentrations of environmental chemicals by integrating ADME parameters, molecular descriptors, and graph-based molecular features. The model outperforms conventional and single-mod…

A multi-task framework combines transformer-based lesion analysis with radiomics for automated diabetic retinopathy screening and severity grading across multiple clinical centers. SegFormer and Deformable DETR achieve strong lesion segmentation and detection…

A multimodal machine learning study evaluates TabPFN alongside SVM, Random Forest, XGBoost, and logistic regression for distinguishing benign and malignant hypervascular thyroid nodules. Using radiomics and deep learning features extracted from B-mode and Dop…

A physics-based machine learning framework combines molecular dynamics-derived descriptors with TabPFN to predict drug formulation properties from limited experimental data. The approach improves generalization for amorphous solid dispersion stability and ach…

A longitudinal study evaluates machine learning models for predicting negative self-rated oral health using sociodemographic, behavioral, and health data from 3,461 adults in Brazil. TabPFN achieves the best performance among five evaluated models with an AUC…

An interpretable machine learning study predicts schizophrenia using sense of agency features derived from agency ratings, time estimation, and intentional binding. Random Forest and TabPFN achieve the strongest individual performance, and their ensemble, Sch…

A TabPFN-based machine learning approach predicts wildfire propagation types in Canadian conifer forests, distinguishing between surface, passive crown, and active crown fires. TabPFN outperforms other machine learning methods in independent evaluations, whil…