Research map / Reinforcement learning / Policy gradient and actor-critic
Hessian Aided Policy Gradient
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 |
|---|---|---|---|
| Add Hessian-aided gradient difference estimator (policy Hessian-vector products along iterate path) to policy gradient | REINFORCE-style policy gradient · theoretical analysis, epsilon-FOSP | trajectory complexity: O(1/eps^4) to O(1/eps^3) | Helped |
| Use variance-reduced SVRPG-style estimator with importance weights instead of Hessian-aided differences | SVRPG · theoretical analysis, epsilon-FOSP | trajectory complexity | No clear effect |
| Replace REINFORCE with HAPG (Hessian-aided estimator) on continuous control tasks | REINFORCE · CartPole, Swimmer, Walker2d, Reacher, Humanoid, HumanoidStandup (system probes) | average episode return vs system probes | Mixed |
Closest papers by meaning
- Revisiting Scalable Hessian Diagonal Approximations for Applications in Reinforcement Learning (ICML 2024 · similarity 0.90)
- Momentum-Based Policy Gradient Methods (ICML 2020 · similarity 0.90)
- Stochastic Policy Gradient Methods: Improved Sample Complexity for Fisher-non-degenerate Policies (ICML 2023 · similarity 0.90)
- Sample Efficient Policy Gradient Methods with Recursive Variance Reduction (ICLR 2020 · similarity 0.89)
- Reusing Trajectories in Policy Gradients Enables Fast Convergence (ICML 2026 · similarity 0.88)
- Natural Policy Gradient as Doubly Smoothed Policy Iteration: A Bellman-Operator Framework (NeurIPS 2026 · similarity 0.87)
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.