TY  - RPRT
TI  - Causal2Vec: Improving Decoder-only LLMs as Embedding Models through a Contextual Token
AU  - Ailiang Lin
AU  - Zhuoyun Li
AU  - Yusong Wang
AU  - Kotaro Funakoshi
AU  - Manabu Okumura
PY  - 2026
DO  - 10.18653/v1/2026.acl-long.1042
UR  - https://arxiv.org/abs/2507.23386
ID  - 2507.23386
ER  -