Research map / Reinforcement learning / Visual and representation RL
How Private Is Your RL Policy? An Inverse RL Based Analysis Framework
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 |
|---|---|---|---|
| Vary the privacy budget across the tested values | Private RL policies evaluated with PRIL · FrozenLake 5x5 and 10x10 | Reward reconstruction distance | No clear effect |
| Use DQN policies instead of PPO or VI policies | Private RL policies evaluated with PRIL · FrozenLake | Reward reconstruction distance | Helped |
| Use VI policies instead of DQN or PPO policies | Private RL policies evaluated with PRIL · FrozenLake | Reward reconstruction distance | Hurt |
| Use DQN-DP-FN rather than the other evaluated private methods | Private RL policies · FrozenLake 5x5 and 10x10 | Test-time return | Hurt |
| Compare private policies with their non-private versions | Private RL policies · FrozenLake | Reward reconstruction distance | No clear effect |
Closest papers by meaning
- On the Sample Complexity of Differentially Private Policy Optimization (NeurIPS 2025 · similarity 0.88)
- Differentially Private Reinforcement Learning with Self-Play (NeurIPS 2024 · similarity 0.88)
- Towards Differentially Private Reinforcement Learning with General Function Approximation (NeurIPS 2026 · similarity 0.88)
- Stealthy Imitation: Reward-guided Environment-free Policy Stealing (ICML 2024 · similarity 0.86)
- A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO (ICML 2025 · similarity 0.86)
- Differentially Private Sparse Reward Estimation with Preference Feedback (NeurIPS 2026 · 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.