Research map / Reinforcement learning / Policy gradient and actor-critic
Predictive CVaR Q-learning
Ju-Hyun Kim, Seungki Min
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 predictive tail value/probability recursion instead of terminal CVaR-Q formulation | PCVaR-Q · Sequential decision tree (q=0.1) | learning stability | Improved |
| optimize CVaR with PCVaR-Q instead of risk-neutral policy | Q-learning · Sequential decision tree (q=0.1) | CVaR of total reward: 1.96 -> 2.45 | Improved |
| optimize CVaR with PCVaR-Q instead of risk-neutral policy | Q-learning · Stochastic grid-world (q=0.1) | CVaR of total reward: −58.37 -> −55.84 | Improved |
| use PCVaR-Q instead of CVaR-Q baseline | Q-learning · Stochastic grid-world (q=0.1) | convergence stability | Improved |
| add risk-level exploration (two-way) without pretraining (ablation) | PCVaR-Q · Grid-world policy visualization | learned path quality | No clear effect |
| add pretraining of parameters with risk-neutral trajectories (ablation) | PCVaR-Q · Grid-world policy visualization | learned path quality | Improved |
| add two-way exploration on top of pretraining (ablation) | PCVaR-Q · Grid-world policy visualization | convergence speed and consistency | Improved |
Closest papers by meaning
- Risk-Sensitive Reward-Free Reinforcement Learning with CVaR (ICML 2024 · similarity 0.90)
- Provably Efficient CVaR RL in Low-rank MDPs (ICLR 2024 · similarity 0.89)
- Provably Efficient Iterated CVaR Reinforcement Learning with Function Approximation and Human Feedback (ICLR 2024 · similarity 0.89)
- A Reductions Approach to Risk-Sensitive Reinforcement Learning with Optimized Certainty Equivalents (ICML 2025 · similarity 0.87)
- Long-Horizon Q-Learning: Accurate Value Learning via n-Step Inequalities (NeurIPS 2026 · similarity 0.87)
- Q-Distribution guided Q-learning for offline reinforcement learning: Uncertainty penalized Q-value via consistency model (NeurIPS 2024 · 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.
