TY  - RPRT
TI  - Text-to-Image Models Need Less from Text Encoders Than You Think
AU  - Nurit Spingarn
AU  - Noa Cohen
AU  - Tamar Rott Shaham
AU  - Tomer Michaeli
PY  - 2026
UR  - https://arxiv.org/abs/2606.03715
ID  - 2606.03715
ER  -