This repo is the official implementation of ICLR 2025 paper "Learning Fine-Grained Representations through Textual Token Disentanglement in Composed Video Retrieval". With the explosive growth of ...