Reproduce “Tabular Foundation Models in Financial Predictions” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ed deHaan, Allen H. Huang
TabFM predicts whether a firm’s net income will rise next year by using labeled historical rows as context, without task-specific training. The authors test it on 6,569 firm-years from 2023 and 2024, comparing its ranking performance with fine-tuned Llama and tree models. TabFM’s pooled AUC is 69.53%, above Llama’s 67.88%; this evidence is limited to one directional-forecast task.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.