3 ms·LLM2CLIP: Powerful Language Model Unlocks Richer Visual Representation1 points by zerojames 2y ago