Hugging Face's logo
AI-ISL 's Collections
BenchPreS: A Benchmark for Context-Aware Personalized Prefer
Rainbow-Padding
R-TOFU: Unlearning in Large Reasoning Models
Model with SAFEPATH
DUSK: Do not Unlearn Shared Knowledge

BenchPreS: A Benchmark for Context-Aware Personalized Prefer

BenchPreS evaluates whether persistent-memory LLMs can selectively apply or suppress stored user preferences based on communicative context.

Upvote
-