Reproduce “TabAbax: A Tabular Foundation Model Technical Report” 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
Zhixiong Cheng, Xiaoyun Gao, Shiyuan Han, Jingjing Suo, Zhengli Wang
TabAbax trains on varied synthetic table tasks and uses a two-pass feature encoder while retaining the TabICLv2 in-context Transformer. On TabArena’s 13-dataset regression suite, its prior-only variant, with the baseline architecture unchanged, scores 1,916 Elo, 228 above TabICLv2; the full model scores 2,024 and ranks seventh of 99 methods. This supports prior changes under this comparison.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.