Leaky Language Models: Stealing Architecture and Inference Optimizations

(arxiv.org)

2 points | by sbulaev 8 hours ago ago