SPREAD: Subspace Representation Distillation for Lifelong Imitation Learning

ArXi:2603.08763v1 Announce Type: new A key challenge in lifelong imitation learning (LIL) is enabling agents to acquire new skills from expert nstrations while retaining prior knowledge. This requires preserving the low-dimensional manifolds and geometric structures that underlie task representations across sequential learning. Existing distillation methods, which rely on L2-norm feature matching in raw feature space, are sensitive to noise and high-dimensional variability, often failing to preserve intrinsic task manifolds. To address this, we.