Research map / Image generation / Editing and style transfer
Counterfactual Generation with Identifiability Guarantees
Hanqi Yan, Lingjing Kong, Lin Gui, Yuejie Chi, Eric Xing, Yulan He, Kun Zhang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add independent content-influence module r_c (Indep) (ablation) | CPVAE · IMDB | accuracy: 20.15 -> 45.00 | Improved |
| add independent content-influence module r_c (Indep) (ablation) | CPVAE · IMDB | BLEU: 49.82 -> 30.88 | Hurt |
| add independent content-influence module r_c (Indep) (ablation) | CPVAE · Yelp | G-score: 16.84 -> 21.24 | Improved |
| add style flow r_s modeling content-style dependency (CausalDep) over Indep (ablation) | CPVAE · IMDB | BLEU: 30.88 -> 39.63 | Improved |
| add style flow r_s (CausalDep) over Indep (ablation) | CPVAE · Yelp | G-score: 21.24 -> 25.98 | Improved |
| add L_sparsity style sparsity regularization to CausalDep (ablation) | CPVAE · IMDB | BLEU: 39.63 -> 56.59 | Improved |
| add L_sparsity to CausalDep (ablation) | CPVAE · Yelp | G-score: 25.98 -> 14.59 | Hurt |
| add L_partial to CausalDep+L_sparsity (ablation) | CPVAE · IMDB | accuracy: 21.55 -> 30.18 | Improved |
| add L_partial to CausalDep+L_sparsity (ablation) | CPVAE · Yelp | accuracy: 13.20 -> 33.70 | Improved |
| add L_c-mask (full MATTE) over L_partial (ablation) | CPVAE · IMDB | accuracy: 30.18 -> 32.43 | Improved |
| full MATTE vs CausalDep (ablation) | MATTE · Yahoo | G-score: 21.39 -> 29.01 | Improved |
| MATTE (causal modules plus sparsity) over CPVAE backbone | CPVAE · IMDB | G-score: 20.01 -> 25.92 | Improved |
Closest papers by meaning
- Towards Characterizing Domain Counterfactuals for Invertible Latent Causal Models (ICLR 2024 · similarity 0.86)
- Gumbel Counterfactual Generation From Language Models (ICLR 2025 · similarity 0.86)
- Content-Style Learning from Unaligned Domains: Identifiability under Unknown Latent Dimensions (ICLR 2025 · similarity 0.86)
- Counterfactual Generative Modeling with Variational Causal Inference (ICLR 2025 · similarity 0.85)
- Benchmarking Counterfactual Image Generation (NeurIPS 2024 · similarity 0.85)
- Counterfactual Image Editing (ICML 2024 · similarity 0.85)
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.
