| Title: | Bridging the question–answer gap in retrieval-augmented generation : hypothetical prompt embeddings |
|---|
| Authors: | ID Vake, Domen (Author)
ID Vičič, Jernej (Author)
ID Tošić, Aleksandar (Author) |
|---|
| Files: |  RAZ_Vake_Domen_2025.pdf (1,41 MB)
MD5: A642E181DB07BC4D6B4A0B0058B6E8EE
 https://ieeexplore.ieee.org/document/11080443
|
|---|
| Language: | English |
|---|
| Work type: | Article |
|---|
| Typology: | 1.01 - Original Scientific Article |
|---|
| Organization: | FAMNIT - Faculty of Mathematics, Science and Information Technologies
|
|---|
| Abstract: | Retrieval-Augmented Generation (RAG) systems synergize retrieval mechanisms with generative language models to enhance the accuracy and relevance of responses. However, bridging the style gap between user queries and relevant information in document text remains a persistent challenge in retrieval-augmented systems, often addressed by runtime solutions (e.g., Hypothetical Document Embeddings (HyDE)) that attempt to improve alignment but introduce extra computational overhead at query time. To address these challenges, we propose Hypothetical Prompt Embeddings (HyPE), a framework that shifts the generation of hypothetical content from query time to the indexing phase. By precomputing multiple hypothetical prompts for each data chunk and embedding the chunk in place of the prompt, HyPE transforms retrieval into a question-question matching task, bypassing the need for runtime synthetic answer generation. This approach does not introduce latency but also strengthens the alignment between queries and relevant context. Our experimental results on six common datasets show that HyPE can improve retrieval context precision by up to 42 percentage points and claim recall by up to 45 percentage points, compared to standard approaches, while remaining compatible with re-ranking, multi-vector retrieval, query decomposition, and other RAG advancements. |
|---|
| Keywords: | LLM, hypothetical prompt embedding, Retrieval-Augmented Generation (RAG) |
|---|
| Publication date: | 15.07.2025 |
|---|
| Year of publishing: | 2025 |
|---|
| Number of pages: | str. 129952-129961 |
|---|
| Numbering: | Vol. 13 |
|---|
| PID: | 20.500.12556/RUP-21512  |
|---|
| UDC: | 004.8 |
|---|
| ISSN on article: | 2169-3536 |
|---|
| DOI: | 10.1109/ACCESS.2025.3589499 |
|---|
| COBISS.SI-ID: | 244701955 |
|---|
| Publication date in RUP: | 04.08.2025 |
|---|
| Views: | 518 |
|---|
| Downloads: | 4 |
|---|
| Metadata: |    |
|---|
|
:
|
Copy citation |
|---|
| | | | Average score: | (0 votes) |
|---|
| Your score: | Voting is allowed only for logged in users. |
|---|
| Share: |  |
|---|
Hover the mouse pointer over a document title to show the abstract or click
on the title to get all document metadata. |
