Zero-Shot Bayesian Optimization with TabPFN: Competitive with State-of-the-Art without Per-Task Training
It is shown that TabPFN v2 (Hollmann et al., 2025), a pretrained tabular foundation model never trained on Bayesian optimization data, can serve as a drop-in zero-shot BO surrogate, eliminating the per-task fitting step.
Theodore Rogers, Srividya Ponnada
· 0 citations