closerG/ppu-bench
PPU-Bench: Real-World Multimodal Benchmark for Personalized Partial Unlearning PPU-Bench is a real-world multimodal benchmark designed to evaluate personalized partial unlearning in vision-language models. It supports multiple unlearning settings and provides training/evaluation data for different VLM backbones.
0568
No commit history came back for main. The revision may not exist, or the source declined the request.
