Research map / Probabilistic and scientific modeling / Variational and generative sampling
Weakly Supervised Representation Learning with Sparse Perturbations
Kartik Ahuja, Jason Hartford, Yoshua Bengio
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 |
|---|---|---|---|
| apply all d perturbations per example instead of one (ablation) | one-sparse encoder · Synthetic d=6 Normal | MCC: 0.99 -> 0.99 | No clear effect |
| select blocks from all pairs (d choose 2) instead of d adjacent pairs (ablation) | overlapping-perturbation encoder · Image balls d=4 | MCC: 0.710 -> 0.864 | Improved |
| select blocks from all pairs instead of d adjacent pairs (ablation) | overlapping-perturbation encoder · Image 10 balls d=20 | MCC: 0.930 -> 0.969 | Improved |
| increase number of balls (latent dimension) | overlapping-perturbation encoder · Image balls, d adjacent blocks | MCC: 0.710 -> 0.866 | Improved |
| use two-sparse blocks instead of one-sparse perturbations | sparse-perturbation encoder · Image balls d=4 | MCC: 0.994 -> 0.710 | Hurt |
Closest papers by meaning
- Object centric architectures enable efficient causal representation learning (ICLR 2024 · similarity 0.87)
- Towards Principled Representation Learning from Videos for Reinforcement Learning (ICLR 2024 · similarity 0.86)
- A Sparsity Principle for Partially Observable Causal Representation Learning (ICML 2024 · similarity 0.86)
- Causal Representation Learning Made Identifiable by Grouping of Observational Variables (ICML 2024 · similarity 0.86)
- Multi-View Causal Representation Learning with Partial Observability (ICLR 2024 · similarity 0.86)
- Learning Linear Causal Representations from General Environments: Identifiability and Intrinsic Ambiguity (NeurIPS 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.
