MOCA, a two-module transformer for causal effect estimation with one-way cross-attention and stop-gradient feedback, is at best competitive with existing methods in most simulations, and the theory promised in the abstract is absent from the paper.
Linear complexity self-attention with 3rd order polynomials
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2026 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
MOCA: A Transformer-based Modular Causal Inference Framework with One-way Cross-attention and Cutting Feedback
MOCA, a two-module transformer for causal effect estimation with one-way cross-attention and stop-gradient feedback, is at best competitive with existing methods in most simulations, and the theory promised in the abstract is absent from the paper.