Dataset distillation can be formulated as a bi-level meta-learning problem where the outer loop optimizes the metadataset and the inner loop trains a model on the distilled data.

Foundation models [6] such as CLIP [47] have been extensively utilized on diverse domains via pretrain-finetune approaches.

While training with static demonstrations has shown some promise, we show that such methods fall short for controlling real GUIs due to their failure to deal with real world stochasticity and non-stationarity not captured in static observational data.

These selections serve as subgoals that indicate subtasks and guide policy.