From Failures to Learnable Worlds: Policy-Conditioned Environment Generation for Navigation Agents
WEDGE generates navigation training environments from a policy's validated failure modes, beating random and hard-example baselines.
1 stories, newest first · one card per story
WEDGE generates navigation training environments from a policy's validated failure modes, beating random and hard-example baselines.
Today 1.6K readers · 2.5K views · 778 reads · all time 189K views