Research map / Reinforcement learning / Visual and representation RL
Policy-Independent Behavioral Metric-Based Representation for Deep Reinforcement Learning
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 policy-independent conservative state-action discrepancy instead of policy-dependent metrics | SAC with Q²-learning · DeepMind Control Suite, clean background | IQM normalized score | Helped |
| Use policy-independent conservative state-action discrepancy instead of policy-dependent metrics | SAC with Q²-learning · DeepMind Control Suite, clutter background | IQM normalized score | Helped |
| Add MICo representation learning | SAC · DeepMind Control Suite, clean background | IQM normalized score | Hurt |
| Add MICo representation learning | SAC · DeepMind Control Suite, clutter background | normalized score | Mixed |
| Add DBC representation learning | SAC · DeepMind Control Suite, clean and clutter backgrounds | IQM normalized score | Hurt |
| Add DBC-normed representation learning | SAC · DeepMind Control Suite, clean and clutter backgrounds | IQM normalized score | Hurt |
| Use gradients from the conservative state-action discrepancy critic rather than the discrepancy actor to update the encoder (ablation) | Q²-learning · DeepMind Control Suite, clean background | IQM normalized score | Helped |
| Extend MICo to state-action similarity (MICo-SA) (ablation) | MICo · DeepMind Control Suite, clean background | IQM normalized score | No clear effect |
Closest papers by meaning
- BeigeMaps: Behavioral Eigenmaps for Reinforcement Learning from Images (ICML 2024 · similarity 0.88)
- METRA: Scalable Unsupervised RL with Metric-Aware Abstraction (ICLR 2024 · similarity 0.87)
- Representation Learning Enables Scalable Multitask Deep Reinforcement Learning (NeurIPS 2026 · similarity 0.87)
- Bridging State and History Representations: Understanding Self-Predictive RL (ICLR 2024 · similarity 0.86)
- State Chrono Representation for Enhancing Generalization in Reinforcement Learning (NeurIPS 2024 · similarity 0.86)
- Geometry of Neural Reinforcement Learning in Continuous State and Action Spaces (ICLR 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.