AI Paper Atlas

Research map / Reinforcement learning / Imitation learning

Inverse Reinforcement Learning from a Gradient-based Learner

Giorgia Ramponi, Gianluca Drappo, Marcello Restelli

NeurIPS 2020 · Reinforcement learning › Imitation learning · approach: Reward learning from demonstrations · 3 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use LOGEL instead of LfL to infer reward weights from G(PO)MDP learningGridWorld · GridWorld, G(PO)MDP learnerreward-weight recoveryImproved
Use LOGEL instead of LfL to infer reward weights from Q-learningGridWorld · GridWorld, Q-learning learnerlearning steps needed for reward-weight recoveryImproved
Use LOGEL instead of LfL to infer reward weights from SVIGridWorld · GridWorld, SVI learnerlearning steps needed for reward-weight recoveryImproved
Use LOGEL instead of LfL to infer reward weights from PPO learningMuJoCo Hopper · Hoppereffectiveness of recovered rewardImproved
Use LOGEL to infer reward weights from PPO learningMuJoCo Reacher · Reacherlearning speedHurt
Estimate gradients from more trajectoriesLOGEL · GridWorld, G(PO)MDP learnerreward-weight recoveryImproved
Observe more policy-improvement steps when gradient estimates are noisyLOGEL · GridWorld, G(PO)MDP learnerreward-weight recoveryImproved
Apply LOGEL to learners using algorithms other than gradient-based onesLOGEL · GridWorldreward-weight recoveryImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy