Reproducible Vision-Language Models Meet Concepts Out of Pre-Training | Read Paper on Bytez