AI Paper Atlas

Research map / Reinforcement learning / Imitation learning

Robust Imitation of a Few Demonstrations with a Backwards Model

Jung Yeon Park, Lawson L.S. Wong

NeurIPS 2022 · Reinforcement learning › Imitation learning · approach: Offline imitation from imperfect data · 4 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
train the policy on backward-model traces as well as demonstrationsBC · Fetch Pushrobustness (%): 12.1 ± 0.3 -> 14.6 ± 0.6Improved
train the policy on backward-model traces as well as demonstrationsBC · Fetch PickAndPlacerobustness (%): 4.1 ± 0.1 -> 17.5 ± 0.9Improved
train the policy on backward-model traces as well as demonstrationsBC · Point mazesrobustness (%): 49.0 ± 1.9 -> 47.8 ± 3.5; 36.8 ± 3.4 -> 38.6 ± 3.4; 33.7 ± 1.5 -> 38.9 ± 2.3Mixed
train the policy on backward-model traces as well as demonstrationsBC · Ant mazesrobustness (%): 63.0 ± 1.0 -> 64.8 ± 1.5; 33.2 ± 0.9 -> 29.1 ± 0.8; 21.7 ± 0.6 -> 17.6 ± 0.5Mixed
train the policy on backward-model traces as well as demonstrationsBC · Adroit Relocaterobustness (%): 7.9 ± 0.7 -> 13.3 ± 1.0Improved
replace the backwards dynamics model with a forward dynamics model (ablation)BMIL · Fetch Pushrobustness (%): 12.1 ± 0.3 -> 12.4 ± 0.6No clear effect
replace the backwards dynamics model with a forward dynamics model (ablation)BMIL · Fetch PickAndPlacerobustness (%): 4.1 ± 0.1 -> 4.1 ± 0.2No clear effect
perturb the first generated action during trace generation (ablation)BMIL · Fetch PickAndPlacerobustness (%)Improved
increase the number of demonstrations (ablation)BC and BMIL · Fetch PickAndPlacerobustness (%)Improved
increase BC gradient steps to match or exceed BMIL compute (ablation)BC · Fetch domainsrobustness (%)Hurt
train the backwards model before training the policy (ablation)BMIL · Fetch Pushrobustness (%): 14.6 ± 0.6 -> 13.2 ± 1.0No clear effect
train the backwards model before training the policy (ablation)BMIL · Fetch PickAndPlacerobustness (%): 17.5 ± 0.9 -> 19.9 ± 2.4No clear effect

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