AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Trust Your $\nabla$: Gradient-based Intervention Targeting for Causal Discovery

Mateusz Olko, Michał Zając, Aleksandra Nowak, Nino Scherrer, Yashas Annadani, Stefan Bauer, Łukasz Kuciński, Piotr Miłoś

NeurIPS 2023 · Probabilistic and scientific modeling › Causal discovery · approach: Interventional causal discovery · 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
use GIT targeting instead of RandomENCO · 24 setups, 12 graphs x 2 regimesmean AUSHD best or comparable count: 12 -> 18Improved
use GIT targeting instead of RandomENCO · 24 setupsmean SHD best or comparable count: 22 -> 17Mixed
use GIT instead of AITENCO · 24 setupsmean AUSHD best or comparable count: 6 -> 18Improved
use GIT instead of CBEDENCO · 24 setupsmean AUSHD best or comparable count: 6 -> 18Improved
restrict to low data regime N=1056GIT · 12 graphsbetter or comparable to other non-privileged methods: 11 out of 12 graphsImproved
replace model-imagined data with real interventional data (GIT-privileged) (ablation)GIT · 24 setupsmean AUSHD best or comparable count: 18 -> 24Improved
use model-sampled data instead of real data in GIT score, low vs regular regime (ablation)GIT · 24 setupsequivalence with GIT-privileged: 10 cases low data, 5 cases regularMixed
increase interventional batch size from 32 to 1024GIT · synthetic 25-node graphsAUSHD vs baselines and final SHD: SHD 0 on all graphs; collider 5.0 vs Random 9.9Improved
increase interventional batch size from 32 to 1024CBED and AIT · synthetic 25-node graphsAUSHD vs Random: collider 9.9 -> 11.4Hurt
increase Monte-Carlo sample size for GIT score (ablation)GIT · ENCOperformanceNo clear effect
use GIT on graphs with singleton Markov equivalence class (cancer, earthquake)GIT · real-world graphsAUSHDNo clear effect
GIT vs GIT-privileged intervention choicesGIT · real-world graphsSpearman correlation of target choices: 73%Improved

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