Simulation study finds TabPFN-based causal estimators biased and slow while CausalPFN shows poor credible-interval coverage in real-world-like scenarios.
Metalearners for estimating heterogeneous treatment effects using machine learningProceedings of the National Academy of Sci- ences.2019;116:4156–4165
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.AP 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Prior-Data Fitted Networks for Causal Inference: a Simulation Study with Real-World Scenarios
Simulation study finds TabPFN-based causal estimators biased and slow while CausalPFN shows poor credible-interval coverage in real-world-like scenarios.