@misc{indiciaec300c434c1d0,
  title = {Causal2Vec: Improving Decoder-only LLMs as Embedding Models through a Contextual Token},
  author = {Ailiang Lin and Zhuoyun Li and Yusong Wang and Kotaro Funakoshi and Manabu Okumura},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1042},
  url = {https://arxiv.org/abs/2507.23386},
  note = {Source identifier: 2507.23386}
}