AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

VER: Scaling On-Policy RL Leads to the Emergence of Navigation in Embodied Rearrangement

Erik Wijmans, Irfan Essa, Dhruv Batra

NeurIPS 2022 · Reinforcement learning › Visual and representation RL · approach: Skill learning and diffusion policies · 6 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 variable experience rollouts instead of equal-length rollouts (ablation)VER versus NoVER · Open Fridge, 8 GPUsMean throughput (SPS): 2216 ±60 -> 2861 ±21Improved
use VER instead of DD-PPOdistributed on-policy RL · PointNav, HM3DThroughput (SPS): 3065 SPS to 5325 SPSImproved
use VER instead of DD-PPOdistributed on-policy RL · ObjectNav, MP3DThroughput (SPS): 1021 to 2019Improved
allow navigation actions during skill training (ablation)TP-SRL skills · Tidy HouseSuccess: 2% Success to 32%Improved
train Pick and Place skills with navigation actions but omit a separate navigation skillTP-SRL(NoNav) · NavPickSuccess: 50%Improved
train Pick and Place skills with navigation actions but omit a separate navigation skillTP-SRL(NoNav) · NavPickNavPlaceSuccess: 20%Improved
add navigation actions to skills (ablation)TP-SRL skills · Prepare GroceriesSuccess: 0% to 5%Improved
add navigation actions to skills (ablation)TP-SRL skills · Set Table, first pick+placeSuccess: 0% to 7.5%Improved
allow skills to navigate for recoveryTP-SRL(NoNav) · Prepare Groceries, first interactionSuccess: -34% SuccessHurt
allow skills to navigate for recoveryTP-SRL(NoNav) · Set Table, first interactionSuccess: -45% SuccessHurt
retain stale data from preempted rollouts for extra PPO epochs (ablation)VER · Open Fridge, 8 GPUsSample efficiencyImproved
collect more experience from faster-simulating environments (ablation)VER · ObjectNav, MP3DSample efficiencyImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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