Research map / Reinforcement learning / Policy gradient and actor-critic
Diverse Exploration via Conjugate Policies for Policy Gradient Methods
Andrew Cohen, Xingye Qiao, Lei Yu, Elliot Way, Xiangrong Tong
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 conjugate perturbations rather than random perturbations | TRPO · Hopper | total pairwise KL divergence: 53.5 vs. 38.1 | Improved |
| Use conjugate perturbations rather than random perturbations | TRPO · Walker | total pairwise KL divergence: 82.7 vs. 77.6 | Improved |
| Use conjugate perturbations rather than random perturbations | TRPO · HalfCheetah | total pairwise KL divergence: 192.5 vs. 156.1 | Improved |
| Use conjugate perturbations rather than random perturbations | TRPO · Hopper | average performance of all behavior policies | Improved |
| Use conjugate perturbations rather than random perturbations | TRPO · Walker | average performance of all behavior policies | Improved |
| Use conjugate perturbations rather than random perturbations | TRPO · HalfCheetah | average performance of all behavior policies | Improved |
| Increase the number of perturbed policies (ablation) | DE on TRPO · Hopper | average performance | Improved |
| Increase the number of perturbed policies (ablation) | DE on TRPO · Hopper | performance variance | Improved |
Closest papers by meaning
- Diverse Exploration for Fast and Safe Policy Improvement (AAAI 2018 · similarity 0.89)
- Diversity-Driven Exploration Strategy for Deep Reinforcement Learning (NeurIPS 2018 · similarity 0.89)
- DGPO: Discovering Multiple Strategies with Diversity-Guided Policy Optimization (AAAI 2024 · similarity 0.87)
- Learning Optimal Deterministic Policies with Stochastic Policy Gradients (ICML 2024 · similarity 0.86)
- Does This Gradient Spark Joy? (NeurIPS 2026 · similarity 0.86)
- Occupancy-based Policy Gradient: Estimation, Convergence, and Optimality (NeurIPS 2024 · 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.
