AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Amortized Inference for Causal Structure Learning

Lars Lorch, Scott Sussex, Jonas Rothfuss, Andreas Krause, Bernhard Schölkopf

NeurIPS 2022 · Probabilistic and scientific modeling › Causal discovery · approach: Constraint-based structure learning with latents · 6 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use AVICI instead of the strongest observational baselineCausal structure learning · Linear, observationalSID: 215.6 ± 35.0 -> 145.6 ± 21.5Improved
Use AVICI instead of the strongest observational baselineCausal structure learning · Rff, observationalF1: 0.476 ± 0.05 -> 0.618 ± 0.06Improved
Use AVICI instead of the strongest observational baselineCausal structure learning · Grn, observationalSID: 588.7 ± 36.6 -> 641.7 ± 34.7Hurt
Use AVICI instead of the strongest interventional baselineCausal structure learning · Linear, interventionalF1: 0.736 ± 0.03 -> 0.819 ± 0.02Improved
Use AVICI instead of the strongest interventional baselineCausal structure learning · Rff, interventionalSID: 282.8 ± 46.3 -> 192.7 ± 44.8Improved
Use AVICI instead of the strongest interventional baselineCausal structure learning · Grn, interventionalF1: 0.092 ± 0.01 -> 0.338 ± 0.06Improved
Increase network depth from 1 to 8 blocks (ablation)AVICI · Rff in-distributionSID: 267.2 ± 22.0 -> 65.2 ± 8.4Improved
Increase network depth from 2 to 8 blocks (ablation)AVICI · Rff in-distributionSID: 195.9 ± 18.5 -> 65.2 ± 8.4Improved
Increase network depth from 4 to 8 blocks (ablation)AVICI · Rff in-distributionSID: 116.6 ± 13.1 -> 65.2 ± 8.4Improved
Add attention over observation axis n (ablation)AVICI · Rff in-distributionSID: 351.5 ± 27.9 -> 65.2 ± 8.4Improved
Add attention over variable axis d (ablation)AVICI · Rff in-distributionSID: 416.8 ± 29.6 -> 65.2 ± 8.4Improved
Replace the inner-product theta model with a relational-network theta model (ablation)AVICI · RffSID: 72.4 ± 9.2 vs. 65.2 ± 8.4; 225.7 ± 25.2 vs. 221.5 ± 24.7No clear effect

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy