BGE Landmark Embedding: A Chunking-Free Embedding Method For Retrieval Augmented Long-Context Large Language Models.
Kun LuoZheng LiuShitao XiaoKang LiuPublished in: CoRR (2024)
Keyphrases
- language model
- context sensitive
- probabilistic model
- text retrieval
- relevance model
- language models for information retrieval
- information retrieval
- document length
- smoothing methods
- language modeling
- retrieval model
- information retrieval systems
- retrieval method
- word segmentation
- passage retrieval
- translation model
- error rate