Research map / Label-efficient and robust learning / Recognition and long-tail
Disentangling Influence: Using disentangled representations to audit model predictions
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use disentangled influence audits with a handcrafted disentangled representation instead of direct SHAP | x+y regression model · Synthetic x+y | feature influence identification | Helped |
| Use learned adversarial-autoencoder disentanglement instead of handcrafted disentanglement (ablation) | Disentangled influence audits · Synthetic x+y | accuracy of influence estimates | Hurt |
| Audit indirect influence rather than direct influence | Adult Income classifier · Adult Income | sex influence | Helped |
| Audit indirect influence rather than direct influence | Adult Income classifier · Adult Income | nationality influence | Helped |
| Use disentangled influence audits instead of BBA | x+y regression model · Synthetic x+y | feature influence discrimination | Helped |
| Use a learned disentangled representation instead of a handcrafted one (ablation) | Disentangled influence audits · Synthetic x+y | variation in influence estimates | Hurt |
| Use disentangled influence audits on dSprites images | Shape classifier · dSprites | latent-factor identification | Helped |
| Use the audit to estimate shape influence on dSprites | Shape classifier · dSprites | shape influence magnitude | Hurt |
Closest papers by meaning
- On Disentangled Representations Learned from Correlated Data (ICML 2021 · similarity 0.87)
- Data-faithful Feature Attribution: Mitigating Unobservable Confounders via Instrumental Variables (NeurIPS 2024 · similarity 0.85)
- Learning Disentangled Representations for CounterFactual Regression (ICLR 2020 · similarity 0.85)
- The Mirrored Influence Hypothesis: Efficient Data Influence Estimation by Harnessing Forward Passes (CVPR 2024 · similarity 0.85)
- Rank-Transformed Dissimilarity Profiles for High-Dimensional Classification (NeurIPS 2026 · similarity 0.84)
- Feature Responsiveness Scores: Model-Agnostic Explanations for Recourse (ICLR 2025 · similarity 0.84)
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.