Statistical Learning
Private Geometric Median Mahdi Haghifam Thomas Steinke Jonathan Ullman
The predominant algorithm for DP convex optimization is DP (stochastic) gradient descent, or DP-(S)GD, for short [SCS13; BST14].
Training Dynamics of Transformers to Recognize Word Co-occurrence via Gradient Flow Analysis
In the literature of studying training dynamics of transformers, several simplifications are commonly adopted such as weight reparameter-ization, attention linearization, special initialization, and lazy regime.
Federated Transformer: Multi-Party Vertical Federated Learning on Practical Fuzzily Linked Data
Federated Learning (FL) is an evolving paradigm that enables multiple parties to collaboratively train models without sharing raw data.
Enhancing Domain Adaptation through Prompt Gradient Alignment
In particular, prior works often generate pseudo-label for unlabeled images and then train the model on those samples.
