AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Steering Visuomotor Policy in Open Worlds via Cross-View Goal Alignment

Shaofei Cai, Zhancun Mu, Anji Liu, Yitao Liang

AAAI 2026 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Imitation and skill learning · 9 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
add target visibility loss to behavior cloning (ablation)ROCKET-2 · Minecraft (3 tasks avg)success rate: 0.65 -> 0.71Improved
add cross-view consistency loss on top of visibility loss (ablation)ROCKET-2 · Minecraft (3 tasks avg)success rate: 0.71 -> 0.94Improved
rely on behavior cloning loss only (ablation)ROCKET-2 · Minecraft (3 tasks avg)success rate: 0.94 -> 0.65Hurt
zero-shot transfer to Unreal with action mapping vs YOLO-Gemma 3ROCKET-2 · ATEC 2025 challengesuccess rate: 26% -> 38%Improved
decouple goal view from agent view (cross-view goals)ROCKET-2 vs ROCKET-1 · Minecraft Interaction Benchmarkinference speed: 3x to 6x speedupImproved

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