Model Releases
DEPT: Document Embedding Preservation Tuning for Unified Query Expansion and Retrieval
arXiv:2608.17632v1 Announce Type: cross Abstract: Large language models (LLMs) can both expand underspecified queries and encode text as dense representations, suggesting a unified model for query exp
arXiv:2608.17632v1 Announce Type: cross Abstract: Large language models (LLMs) can both expand underspecified queries and encode text as dense representations, suggesting a unified model for query expansion and retrieval. Existing systems usually rely on prompted expansions, independently trained modules, or staged optimization, leaving generated expansions only indirectly aligned with the retrieval loss that judges them. We train a single decoder-only LLM end to end, where the same model generates the expansion and encodes both the expanded query and candidate documents. This unified setting creates a moving-target problem: retrieval supervision should improve query-side expansion, but the same update also shifts the document embeddings that serve as retrieval targets. We introduce Document Embedding Preservation Tuning (DEPT), which keeps tuned document embeddings close to cached initial embeddings while allowing retrieval gradients to pass through straight-through decoding into the generator. DEPT converts joint query--document movement into query-side adaptation against approximately stable, whitened document embeddings that support index reuse and online hard-negative mining. Experiments with Qwen3-4B-Instruct-2507 and LLaMA-3.2-3B-Instruct on five datasets in BEIR benchmark show that DEPT improves average retrieval quality over training-free, independently trained, and staged unified baselines, while ablations isolate the effects of preservation, whitening, end-to-end expansion training, and online negatives. Code is available at https://github.com/ILSparkle/DEPT.
Related
- Hybrid privacy-aware semantic search: SVD-truncated document geometry and CKKS-encrypted query reranking under a restricted threat model
- On the Robustness of Multilingual Text Embedding Rankings Across Learning Tasks, Languages, and Benchmark Datasets
- Conan-embedding-v3: Fusing Modality-Specific Models for Omni-Modal Embedding
- Multimodal Hybrid Retrieval-Augmented Generation for Scientific Document Understanding using Open-Source SLMs
Source: arXiv cs.AI | 2026-08-19