Search Results for author:
Found 4 papers, 3 papers with code
Position: LLM Unlearning Benchmarks are Weak Measures of Progress
Unlearning methods have the potential to improve the privacy and safety of large language models (LLMs) by removing sensitive or harmful information post hoc.
Unified Locational Differential Privacy Framework
Aggregating statistics over geographical regions is important for many applications, such as analyzing income, election results, and disease spread.
Guardrail Baselines for Unlearning in LLMs
Recent work has demonstrated that finetuning is a promising approach to 'unlearn' concepts from large language models.
IS-CAM: Integrated Score-CAM for axiomatic-based explanations
Convolutional Neural Networks have been known as black-box models as humans cannot interpret their inner functionalities.
