Research map / Probabilistic and scientific modeling / Molecules and proteins
Counterfactuals uncover the modular structure of deep generative models
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 |
|---|---|---|---|
| Cluster elementary influence maps with NMF rather than k-means | β-VAE on CelebA · CelebA | cluster consistency | Helped |
| Use 3 clusters for influence-map clustering | β-VAE on CelebA · CelebA | cluster consistency: >90% | Helped |
| Use 3 clusters for influence-map clustering | β-VAE on CelebA · CelebA | average cosine similarity between matched templates: .9 | Helped |
| Intervene on channels grouped into a module to hybridize two generated images | β-VAE on CelebA · CelebA | selective feature transfer | Helped |
| Intervene on channels from the same cluster in two successive layers | BEGAN · CelebA | selective feature transfer | Helped |
| Intervene on two successive layers within a Gblock | BigGAN-deep · ImageNet | hybrid image quality | Helped |
| Hybridize features across classes using internal counterfactual interventions | BigGAN-deep · ImageNet | meaningful cross-class hybrid generation | Helped |
| Increase the intervention depth toward layers closest to the output | BigGAN-deep and pretrained classifiers · ImageNet | recognition rate of the most recognized original class | Helped |
Closest papers by meaning
- Counterfactual Image Editing (ICML 2024 · similarity 0.88)
- Counterfactual Generative Modeling with Variational Causal Inference (ICLR 2025 · similarity 0.87)
- Benchmarking Counterfactual Image Generation (NeurIPS 2024 · similarity 0.87)
- V-CECE: Visual Counterfactual Explanations via Conceptual Edits (NeurIPS 2025 · similarity 0.86)
- Rethinking Visual Counterfactual Explanations Through Region Constraint (ICLR 2025 · similarity 0.86)
- Diffusion Counterfactual Generation with Semantic Abduction (ICML 2025 · similarity 0.86)
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.