Research map / Probabilistic and scientific modeling / Treatment effects
Addressing Instrument-Outcome Confounding in Mendelian Randomization through Representation Learning
Shimeng Huang, Matthew R Robinson, Francesco Locatello
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use learned invariant component W-hat as instrument instead of observed Z | 2SLS · AoU GLP1R semi-synthetic MR | bias of ACE: bias up to 0.3 -> near zero | Improved |
| partial out the discrete population label K | 2SLS(Z) and Egger(Z) · AoU GLP1R semi-synthetic MR | bias of ACE | Mixed |
| add partialling out of learned variant component V-hat | 2SLS(W-hat) · AoU GLP1R semi-synthetic MR | estimator variance | Improved |
| over-specify latent dimension to 4 (true p=2) (ablation) | representation learner · Simulated, polynomial degree 3 | bias and variance | Hurt |
| under-specify latent dimension (p-hat<2) (ablation) | representation learner · Simulated, polynomial degree 3 | bias | No clear effect |
| add independence loss (lambda2>0) (ablation) | PO(V-hat)-2SLS(W-hat) · Simulated, Poly 3 and MLP mixing | bias | Improved |
| add independence loss (lambda2>0) (ablation) | 2SLS(W-hat) · Simulated mixings | bias | No clear effect |
Closest papers by meaning
- Identification and Estimation of the Bi-Directional MR with Some Invalid Instruments (NeurIPS 2024 · similarity 0.89)
- Learning Representations of Instruments for Partial Identification of Treatment Effects (ICML 2025 · similarity 0.85)
- Automating the Selection of Proxy Variables of Unmeasured Confounders (ICML 2024 · similarity 0.84)
- Doubly robust identification of treatment effects from multiple environments (ICLR 2025 · similarity 0.83)
- Conditional Instrumental Variable Regression with Representation Learning for Causal Inference (ICLR 2024 · similarity 0.83)
- Training Flexible Models of Genetic Variant Effects from Functional Annotations using Accelerated Linear Algebra (ICML 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
