Tag: foundation-model reinforcement learning

Why it matters: Inside DeepMind XLand: the 3D task universe, XLand 2.0, the Adaptive Agent, and what open-ended learning means for 2026.