作者位次 #2/8ACL 2026Poster· Natural Language Processing
Focusing Condition: Inference-Time Self-Contrastive Steering Elicits Better Conditional Text Embeddings in LLM s
Zifeng Cheng · Lingyun Qian · Zhiwei Jiang · Cong Wang · Yafeng Yin · Fei Shen · Ao Zhou · Qing Gu
摘要
Extracting conditional text embeddings from large language models (LLMs) is a promising paradigm, as it requires neither additional data nor fine-tuning. Existing methods incorporate conditions into prompts to guide LLMs to focus on specific aspects and elicit conditional text embeddings. However...