Humzah Merchant
2 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
Divergence Decoding (DD) is a novel, effective, and inexpensive method that uses auxiliary models to steer LLM logits during inference, enabling the removal of memorized sensitive data without significant utility loss.
Divergence Decoding (DD) is a novel, effective, and inexpensive method that uses auxiliary models to steer LLM logits during inference, enabling the removal of memorized sensitive data without significant utility loss.
Papers
Divergence Decoding: Inference-Time Unlearning via Auxiliary Models
Divergence Decoding (DD) is a novel, effective, and inexpensive method that uses auxiliary models to steer LLM logits during inference, enabling the removal of memorized sensitive data without signifi…