The large-scale propagation of human-human handover research has been hindered by two challenges.

Algorithmic stability is a major theme in learning theory, where seminal results have firmly established its close relationship with generalization. Recent research has further highlighted the intricate interplay between stability and additional properties of interest beyond statistical generalization.

Classically, sketching has been applied to design low-memory algorithms in the streaming setting, when the input is presented to the algorithm as a sequence of updates.

Classically, sketching has been applied to design low-memory algorithms in the streaming setting, when the input is presented to the algorithm as a sequence of updates.

In contrast, in non-private settings, practitioners commonly utilize "adaptive" hyperparameter optimization methods such as Gaussian process-based optimization, which select the next candidate based on information gathered from previous outputs. This substantial contrast between private and non-private hyperparameter optimization underscores a critical concern. In our paper, we introduce DP-HyPO, a pioneering framework for "adaptive"