AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Causal Preference Elicitation

Edwin V. Bonilla, He Zhao, Daniel M. Steinberg

ICML 2026 · Probabilistic and scientific modeling › Causal discovery · approach: Constraint-based structure learning with latents · 0 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
select expert queries by expected information gain (CaPE) instead of random or uncertainty samplingCaPE · Synthetic ER DAG (D=20)entropy, SHD, ETCPImproved
add 40 EIG-selected expert queries to the observational posteriorCaPE · Sachs (observational only)SHD: approximately 14 edgesImproved
add 40 EIG-selected expert queries to the observational posteriorCaPE · Sachs (observational only)orientation F1: more than 0.38Improved
use CaPE instead of random, uncertainty and static-EIG baselinesCaPE · CausalBench K562 50 genesdirected AUPRC and Top-K precision: clearly outperforming all baselines by T >= 100Improved
replace adaptive EIG with static EIG computed once from the initial posterior (STE) (ablation)CaPE · CausalBench K562 50 genesAUPRCHurt
use uncertainty sampling instead of EIG for query selection (ablation)CaPE · CausalBench K562 50 genesAUPRCHurt
initialize particles from a trained DAG-GFlowNet prior instead of bootstrap regressionCaPE · Sachs (observational only)SHD, ETCP, entropyImproved
misspecify the expert oracle (heterogeneous reliability, directional bias, abstention, noisy)CaPE · Synthetic ER DAG (D=20), misspecified expertsETCPNo 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