Research map / Probabilistic and scientific modeling / Treatment effects
Causal Effect Inference with Deep Latent-Variable Models
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Model hidden confounders as a latent variable and infer them from proxies | Causal effect estimation on the synthetic toy data · Synthetic toy data | ATE absolute error | Helped |
| Use CEVAE instead of the direct-proxy baselines | Counterfactual mortality inference · Twins | counterfactual performance | Helped |
| Increase proxy bit-flip noise | CEVAE counterfactual inference · Twins | counterfactual performance | Mixed |
| Use CEVAE rather than direct-proxy methods | Average treatment effect inference · Twins | ATE error | Mixed |
| Use a binary latent variable matching the data-generating confounder rather than a continuous 5-dimensional latent variable (ablation) | CEVAE · Synthetic toy data | ATE absolute error | Helped |
| Use a continuous 5-dimensional latent variable instead of the correctly specified binary latent variable (ablation) | CEVAE · Synthetic toy data | ATE absolute error | Hurt |
| Set CEVAE to zero hidden layers while retaining its latent-space inference (ablation) | CEVAE · Twins counterfactual inference | counterfactual performance | Helped |
| Set CEVAE to zero hidden layers while retaining its latent-space inference (ablation) | CEVAE · Twins | ATE error | Hurt |
| Vary proxy noise from low levels toward completely uninformative proxies | CEVAE · Twins | ATE error | Helped |
Closest papers by meaning
- Causal Inference with Conditional Front-Door Adjustment and Identifiable Variational Autoencoder (ICLR 2024 · similarity 0.90)
- Doubly robust identification of treatment effects from multiple environments (ICLR 2025 · similarity 0.90)
- Automating the Selection of Proxy Variables of Unmeasured Confounders (ICML 2024 · similarity 0.89)
- An Analysis of Causal Effect Estimation using Outcome Invariant Data Augmentation (NeurIPS 2025 · similarity 0.89)
- End-To-End Causal Effect Estimation from Unstructured Natural Language Data (NeurIPS 2024 · similarity 0.89)
- Proximal Causal Inference With Text Data (NeurIPS 2024 · similarity 0.88)
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.