tabicl

a native TabICLv2 implementation for R torch

R package for tabular foundation models: TabPFN, TabICL, and TabFM.

COMP 4332 Project 3 rating prediction using LightGBM, OOF target encoding, stacking, and segment calibration.

Time series forecasting, anomaly detection and tabular prediction in the browser (WebGPU and WASM): TimesFM 2.5, Chronos-Bolt and TabICL, with parity to the official Python libraries.

Let’s Play of Claude Code CLI running AdCP autonomously

Research on continued pretraining of TabPFN and TabICL on credit data for default prediction and loss given default.

Free Claude MCP server for TabICL — tabular ML predictions, explanations and shareable reports

Curated research library on tabular foundation models, with paper summaries, cross-paper synthesis and upstream code snapshots.

Research benchmark comparing tabular foundation models, classical methods and deep learners on credit default and loss given default.

Research on credit-risk synthetic priors for TabICL, covering default prediction and loss given default.

Small Dataset Benchmarks on the Dataset of Datasets UCI++

Serving engine for tabular foundation models (TabICL, TabPFN): fit a table once, keep its encoded context on the GPU, and serve fast repeat predictions to many tenants, with memory-aware admission and multi-GPU placement.

Do tabular foundation models keep their edge under distribution shift? Superconductor Tc prediction and screening (TabICL, TabDPT vs CatBoost)

5th place (private AUC 0.9033) in Kaggle's Rainfall Probability Prediction Challenge: classical stack + TabPFN/TabICL on Kaggle GPUs

Reproducible datasets, computational workflows, and model implementations for externally validated tabular foundation models and explainable concrete compressive strength prediction.

Calibration gate for tabular classifiers (TabICL-first, model-agnostic, offline, CI-ready)

Predict rarely measured soil properties (total N, CEC7) by learning the bounded parameters of known soil relations; spatial benchmark with QRF, TabM and TabICLv2 on public WoSIS data.

Code and data for the survey Tabular Foundation Models: A Systematic Survey.

Out-of-time, calibration and stability evaluation of tabular foundation models (TabPFN, TabICL) on credit default, with pre-registrations, a claim ledger and every recorded run.