Cliff-Learning
This addresses data efficiency in transfer learning for AI applications, but it appears incremental as it focuses on analyzing an observed phenomenon rather than introducing a new method.
The paper investigates cliff-learning, a phenomenon where performance improves faster than a power law rate in low-data transfer learning from foundation models, and finds that the degree of cliff-learning indicates compatibility between algorithm priors and task requirements.
We study the data-scaling of transfer learning from foundation models in the low-downstream-data regime. We observe an intriguing phenomenon which we call cliff-learning. Cliff-learning refers to regions of data-scaling laws where performance improves at a faster than power law rate (i.e. regions of concavity on a log-log scaling plot). We conduct an in-depth investigation of foundation-model cliff-learning and study toy models of the phenomenon. We observe that the degree of cliff-learning reflects the degree of compatibility between the priors of a learning algorithm and the task being learned.