Research map / Probabilistic and scientific modeling / Dynamical systems
Learning Plannable Representations with Causal InfoGAN
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 Causal-InfoGAN instead of K-means | 2D planning · Tunnels | plan feasibility: 12.25% -> 98% | Helped |
| Use Causal-InfoGAN instead of Temporal K-means | 2D planning · Tunnels | plan feasibility: 7.0% -> 98% | Helped |
| Use Causal-InfoGAN instead of Spectral clustering | 2D planning · Tunnels | plan feasibility: 8.75% -> 98% | Helped |
| Use Causal-InfoGAN instead of K-means | 2D planning · Door-key | plan feasibility: 100% -> 98% | Hurt |
| Use Causal-InfoGAN instead of Temporal K-means | 2D planning · Door-key | plan feasibility: 100% -> 98% | Hurt |
| Use Causal-InfoGAN instead of Spectral clustering | 2D planning · Door-key | plan feasibility: 60% -> 98% | Helped |
| Use Causal-InfoGAN instead of K-means | 2D planning · Rescaled door-key | plan feasibility: 0.0% -> 97% | Helped |
| Use Causal-InfoGAN instead of Temporal K-means | 2D planning · Rescaled door-key | plan feasibility: 0.0% -> 97% | Helped |
| Use Causal-InfoGAN instead of Spectral clustering | 2D planning · Rescaled door-key | plan feasibility: 20.0% -> 97% | Helped |
| Add transition self-consistency loss (ablation) | Causal InfoGAN on low-dimensional observations · 2D navigation | training stability | Helped |
| Add transition self-consistency loss (ablation) | Causal InfoGAN on high-dimensional images · Rope manipulation | planning performance | No clear effect |
| Vary binary latent-state dimension from 3 to 5 (ablation) | Causal InfoGAN · 2D key domain | planning performance | No clear effect |
Closest papers by meaning
- Hierarchical Foresight: Self-Supervised Learning of Long-Horizon Tasks via Visual Subgoal Generation (ICLR 2020 · similarity 0.88)
- FLIP: Flow-Centric Generative Planning as General-Purpose Manipulation World Model (ICLR 2025 · similarity 0.88)
- PcLast: Discovering Plannable Continuous Latent States (ICML 2024 · similarity 0.86)
- Visual Planning: Let's Think Only with Images (ICLR 2026 · similarity 0.86)
- Latent Plan Transformer for Trajectory Abstraction: Planning as Latent Space Inference (NeurIPS 2024 · similarity 0.86)
- Differentiable Learning of Lifted Action Schemas for Classical Planning (NeurIPS 2026 · 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.