AI Paper Atlas

Research map / Reinforcement learning / Imitation learning

Where Do You Think You're Going?: Inferring Beliefs about Dynamics from Behavior

Sid Reddy, Anca Dragan, Sergey Levine

NeurIPS 2018 · Reinforcement learning › Imitation learning · approach: Reward learning from demonstrations · 26 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 learned internal dynamics for internal-to-real dynamics transfer assistanceHuman Lunar Lander control · Lunar Lander user studySuccess rate: f(1,11)=109.58, p<0.0001Improved
Use learned internal dynamics for internal-to-real dynamics transfer assistanceHuman Lunar Lander control · Lunar Lander user studyCrash rate: f(1,11)=126.33, p<0.0001Improved
Use learned internal dynamics to assist controlLunar Lander user study · Lunar Lander user studyAverage speed: 30% decreaseImproved
Train on more than one known-reward task (ablation)Inverse soft Q-learning · Grid World NavigationInternal dynamics prediction accuracyImproved
Add the action intent prior restricting predictions to states reachable under real dynamics (ablation)Inverse soft Q-learning · Grid World Navigation with globally scrambled actionsInternal dynamics prediction accuracyImproved
Add the action intent prior restricting predictions to states reachable under real dynamics (ablation)Inverse soft Q-learning · Grid World Navigation with locally scrambled actionsInternal dynamics prediction accuracyMixed
Use constraint sampling and deterministic internal dynamics approximations for continuous states (ablation)Inverse soft Q-learning · 2D Continuous NavigationInternal dynamics parameter errorNo clear effect
Use learned internal dynamics rather than assume internal dynamics equal real dynamicsMaxCausalEnt IRL · Grid World reward learning from misguided demonstrationsTrue reward collected by learned-reward policyImproved
Use inverse soft Q-learning instead of simultaneous reward-and-dynamics estimationSERD baseline · Grid World reward learning from misguided demonstrationsReward learning performanceImproved
Model user behavior using learned internal dynamics instead of real dynamicsMaxCausalEnt IRL · Grid World reward learning from misguided demonstrationsReward learning performanceImproved

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