The 4B version of Think When Needed (TWN), a framework for adaptive reasoning-driven multimodal embeddings. TWN introduces a dual-LoRA architecture on top of Qwen3-VL-4B-Instruct with a learned routing gate that adaptively activates chain-of-thought reasoning.
1@article{zhang2026thinkneeded,
2 title={Think When Needed: Adaptive Reasoning-Driven Multimodal Embeddings with a Dual-LoRA Architecture},
3 author={Longxiang Zhang and Weilong Dai and Guanghao Zhang and Hao Jiang and Pipei Huang},
4 year={2026},
5 eprint={2605.14448},
6 archivePrefix={arXiv},
7 primaryClass={cs.CV},
8 url={https://arxiv.org/abs/2605.14448},
9}
