Research map / Reinforcement learning / Visual and representation RL
Learning World Models with Identifiable Factorization
Yu-Ren Liu, Biwei Huang, Zhengmao Zhu, Honglong Tian, Mingming Gong, Yang Yu, 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 |
|---|---|---|---|
| train SAC on s^r only versus other combinations of the four learned latent categories (s^ar, s^r-bar, s^r+s^a-bar-r, full state) (ablation) | IFactor · RoboDesk with noise distractors | return | Improved |
| include mutual information constraints in the objective (ablation) | IFactor · RoboDesk / DMC | policy return and disentanglement | Improved |
| IFactor vs Denoised MDP on RoboDesk | IFactor · RoboDesk with noise distractors | return | No clear effect |
| IFactor vs baselines across distractor variants | IFactor · DeepMind Control Suite with distractors | return | Improved |
| IFactor vs BetaVAE, FactorVAE, SlowVAE, PCL, TDRL, Denoised MDP on latent recovery | IFactor · Synthetic identifiability | R^2: >0.9 | Improved |
Closest papers by meaning
- Simplifying Latent Dynamics with Softly State-Invariant World Models (NeurIPS 2024 · similarity 0.88)
- Inverting the Bellman Equation: From $Q$-Values to World Models (NeurIPS 2026 · similarity 0.88)
- Beyond Prediction: Steering VLM Agents with Retrospective World Modeling (NeurIPS 2026 · similarity 0.88)
- Modeling Unseen Environments with Language-guided Composable Causal Components in Reinforcement Learning (ICLR 2025 · similarity 0.87)
- World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetry (NeurIPS 2026 · similarity 0.87)
- Policy Regularization on Globally Accessible States in Cross-Dynamics Reinforcement Learning (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.
