AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Causal Intervention in Concept Bottleneck Models

Zhiyu Zhu, Jiayu Zhang, Zhibo Jin, Xinyi Wang, Fang Chen, Przemyslaw Biecek, Jianlong Zhou

ECCV 2026 · Probabilistic and scientific modeling › Causal discovery · approach: Causal representation identifiability · 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
replace Concept Realignment with training-free Causal InterventionUCP policy · CBM benchmarks (first 10 steps)Intervention-AUC@10: 3.90% improvement over CRImproved
use CI instead of other baseline methodsCBM interventions · CBM benchmarks (first 10 steps)Intervention-AUC@10: 8.59% average improvementImproved
use CI at 15 steps instead of CRUCP policy · CBM benchmarks (first 15 steps)Intervention-AUC@15: 9.14% over others, 3.46% over CRImproved
use CI at 20 steps instead of CRUCP policy · CBM benchmarks (first 20 steps)Intervention-AUC@20: 8.85% over others, 2.88% over CRImproved
add CI on top of UCP over all stepsUCP · CBM benchmarks (all steps)Intervention-AUC: 5.07% average gain over mean of competing methodsImproved
apply CI instead of CR with Int-CEM on AwA2Int-CEM · AwA2 individualIntervention-AUC: 8424.78 -> 8421.38No clear effect
raise learning rate from 0.03125 to 0.25 (ablation)CI with CEM · CUB individualIntervention-AUC: 10610.90 -> 10661.75Improved
raise learning rate from 0.03125 to 0.25 (ablation)CI with CBM · CUB groupedIntervention-AUC: 2537.13 vs. 2561.61Hurt
vary learning rate across 0.03125 to 0.25 (ablation)CI · AwA2/CUBIntervention-AUCNo clear effect
use UCP policy instead of COOP policy (ablation)CEM · AwA2 individualIntervention-AUC: 8171.87 -> 8459.34Improved
use UCP policy instead of COOP policy (ablation)Int-CEM · CUB individualIntervention-AUC: 9135.53 -> 11078.61Improved
use UCP instead of EUDTP policy (ablation)CBM · CUB groupedIntervention-AUC: 2506.72 -> 2561.61No 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