Research map / Reinforcement learning / Policy gradient and actor-critic
High-Dimensional Contextual Policy Search with Unknown Context Rewards using Bayesian Optimization
Qing Feng, Ben Letham, Hongzi Mao, Eytan Bakshy
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 |
|---|---|---|---|
| use LCE-A instead of standard GP BO under aggregate rewards | Contextual BO · Contextual Hartmann5D, varying contexts | final reward | Improved |
| observe context-level rewards (ICM, LCE-M) vs aggregate only (LCE-A) (ablation) | Contextual BO · Contextual Hartmann5D | final reward | Improved |
| increase number of sparse contexts with skewed weights (0.01 each) (ablation) | SAC and LCE-A · Contextual Hartmann5D skewed weights | best reward found | No clear effect |
| use LCE-A instead of SAC (ablation) | aggregate-reward contextual BO · Rare-context reward (5 dense, 5 sparse contexts) | reward on sparse contexts | Improved |
| violate Assumption A3 with dynamic context weights wc depending on rewards (ablation) | SAC and LCE-A · 50 contexts dynamic distribution | best reward | No clear effect |
| contextual policy (LCE-A) instead of non-contextual policy | ABR controller · ABR video playback | QoE: 50% increase in QoE | Improved |
| use linear subspace BO (ALEBO, HeSBO, REMBO) instead of LCE-A | ABR policy search · ABR video playback | final best QoE: 25% worse | Hurt |
| use Add-GP-UCB or TuRBO instead of LCE-A | ABR policy search · ABR video playback | final best QoE: 6% less | Hurt |
Closest papers by meaning
- Cost-Aware Near-Optimal Policy Learning (AAAI 2025 · similarity 0.88)
- Zero-Shot Context Generalization in Reinforcement Learning from Few Training Contexts (NeurIPS 2025 · similarity 0.86)
- Reflective Prompted Policy Optimization: Trajectory-Grounded Revision and Salience Bias (NeurIPS 2026 · similarity 0.86)
- XLand-100B: A Large-Scale Multi-Task Dataset for In-Context Reinforcement Learning (ICLR 2025 · similarity 0.85)
- SPOT: Scalable Policy Optimization with Trees for Markov Decision Processes (NeurIPS 2025 · similarity 0.85)
- Integrating Inference and Experimental Design for Contextual Behavioral Model Learning (AAAI 2025 · 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.
