Reproduce “Learning Functional Specialisation for Bimanual Robotic Control” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jingwen Li, Levin Kuhlmann, Jason Friedman, Gideon Kowadlo
The authors test whether controller-specific learning objectives and stable hand-role assignments shape simulated bimanual tracking and holding. Differential and shared objectives are crossed with fixed and balanced hand assignments. Differential-Fixed shows the strongest task-aligned specialization and reaches 81.33% success. Reversal reduces its success by 80.46 percentage points, versus 1.51 for Shared-Balanced; simulation does not establish specialization as the cause.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.