Paper | Bibtex | Code Neural Information Processing Systems (NeurIPS), 2023 We present a new algorithm, Cross-Episodic Curriculum (CEC), to boost the learning efficiency and generalization of Transformer agents. Central to CEC is the placement of cross-episodic experiences into a Transformer’s context, which forms the basis of a curriculum. By sequentially structuring online learning trials and mixed-quality demonstrations, CEC constructs curricula that encapsulate learning progression and proficiency incre
Transformers excel at recognizing patterns, but they struggle when there's limited data for learning agents. For complex tasks, agents either need abundant samples (RL agents) or demonstrations (IL agents), making it challenging in fields like robotics where data is scarce. How can we make the most of the limited data, regardless of their optimality and construction, for more efficient learning? Our insight is that when we examine data across different episodes, useful patterns emerge. For example, an RL ag
In IL settings, human demonstrations vary in quality, but still present patterns of improvement and generally effective manipulation skills among different operators: