AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Unveiling Prior-Data Fitted Networks on Causal Effect Estimation: Pre-Training or Fine-Tuning?

Haotian Wang, Xinpeng Lv, Hao Zou, Yanghao Xiao, Shanzhi Gu, Yang Shi, Yunxin Mao, Yuanxing Zhang, Mingyang Geng, Shaowu Yang, Haoxuan Li, Wenjing Yang and 2 more

ICML 2026 · Probabilistic and scientific modeling › Causal discovery · approach: Latent-variable causal estimation

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
increase number of intervened variables kinterv from 2 to 7MSF fine-tuned TabPFN · Linear SCMMAE: 0.3087 -> 0.3824Hurt
increase number of intervened variables kinterv from 2 to 7MSF fine-tuned CausalPFN · Interaction SCMMAE: 1.3157 -> 1.6487Hurt
increase number of intervened variables kinterv from 2 to 7MSF fine-tuned DoPFN · Interaction SCMMAE: 2.0135 -> 2.5897Hurt
increase number of intervened variables kinterv from 2 to 7MSF fine-tuned TabPFN · Nonlinear SCMMAE: 0.1050 -> 0.1856Hurt
increase TV radius, Lipschitz factor and coverage radiusPWF · Interaction SCMMAE: 0.425 -> 0.535Hurt
increase TV radius, Lipschitz factor and coverage radiusPWF · Linear SCMMAE: 0.134 -> 0.158Hurt
apply point-wise interventional fine-tuning (PWF) on uncovered distribution D1pre-trained PFNs · Linear, Nonlinear, Interaction SCMsMSEImproved
evaluate PWF on distribution D3 far from D1 instead of near D2 (ablation)PWF · synthetic SCMsMSEHurt
evaluate pre-trained PFNs on uncovered instead of covered interventionspre-trained PFNs · LalondeMSE: error > 5Hurt
apply meta-sampling fine-tuning (MSF)pre-trained PFNs · Interaction SCM, |X|=10MAEImproved
increase MSF sampling budget K from 5 to 11 (ablation)MSF · synthetic SCMsMAEImproved

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