Google · ML· Code
LOW · $6,154/entrantTop $/entrant
Google · ML · Build autonomous agent harnesses to solve real-world structural, civil, electrical, and mechanical engineering tasks.
Jan 5
LOW · $6,154/entrant
Code
↗Train an HVAC controller for unseen buildings. The $40,000 cash pool covers $25,000 in leaderboard awards and $15,000 across two paper rounds.
by University of Washington / NeurIPS
Join on Kaggle by December 3 at 23:59 UTC, the cutoff set in the platform. The organizer’s written schedule lists January 3.
Train and benchmark your control policy with the building simulator.
Use JetBrains IDE + AI Assistant — Code & test the project ↗Submit your Kaggle notebook by January 15 at 17:59 UTC. The organizer lists 23:59 UTC that day for final submissions, including the paper track on OpenReview.