{"ID":2840389,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.13018","arxiv_id":"2511.13018","title":"The Final-Stage Bottleneck: A Systematic Dissection of the R-Learner for Network Causal Inference","abstract":"The R-Learner is a powerful, theoretically-grounded framework for estimating heterogeneous treatment effects, prized for its robustness to nuisance model errors. However, its application to network data, where causal heterogeneity is often graph-dependent, presents a critical challenge to its core assumption of a well-specified final-stage model. In this paper, we conduct a large-scale empirical study to systematically dissect the R-Learner framework on graphs. We provide the first rigorous evidence that the primary driver of performance is the inductive bias of the final-stage CATE estimator, an effect that dominates the choice of nuisance models. Our central finding is the quantification of a catastrophic \"representation bottleneck\": we prove with overwhelming statistical significance (p \u003c 0.001) that R-Learners with a graph-blind final stage fail completely (MSE \u003e 4.0), even when paired with powerful GNN nuisance models. Conversely, our proposed end-to-end Graph R-Learner succeeds and significantly outperforms a strong, non-DML GNN T-Learner baseline. Furthermore, we identify and provide a mechanistic explanation for a subtle, topology-dependent \"nuisance bottleneck,\" linking it to GNN over-squashing via a targeted \"Hub-Periphery Trade-off\" analysis. Our findings are validated across diverse synthetic and semi-synthetic benchmarks. We release our code as a reproducible benchmark to facilitate future research on this critical \"final-stage bottleneck.\"","short_abstract":"The R-Learner is a powerful, theoretically-grounded framework for estimating heterogeneous treatment effects, prized for its robustness to nuisance model errors. However, its application to network data, where causal heterogeneity is often graph-dependent, presents a critical challenge to its core assumption of a well-...","url_abs":"https://arxiv.org/abs/2511.13018","url_pdf":"https://arxiv.org/pdf/2511.13018v3","authors":"[\"S Sairam\",\"Sara Girdhar\",\"Shivam Soni\"]","published":"2025-11-17T06:16:04Z","proceeding":"cs.LG","tasks":"[\"cs.LG\"]","methods":"[\"Graph Neural Network\"]","has_code":false}