# Imagine All The Relevance: Scenario-Profiled Indexing with Knowledge Expansion for Dense Retrieval

arXiv: https://arxiv.org/abs/2503.23033 · HTML: https://mlsanity.com/paper/2503.23033/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Mar 2025.

## Built on by (most cited first, 5 of 5)

- [Personalized Reward Modeling for Text-to-Image Generation](https://mlsanity.com/paper/2511.19458/) (2025)
- [RL-Index: Reinforcement Learning for Retrieval Index Reasoning](https://mlsanity.com/paper/2606.16316/) (2026)
- [Offline Reasoning for Efficient Recommendation: LLM-Empowered Persona-Profiled Item Indexing](https://mlsanity.com/paper/2602.21756/) (2026)
- [A Survey of Reasoning-Intensive Retrieval: Progress and Challenges](https://mlsanity.com/paper/2605.00063/) (2026)
- [Self-Evolving Search Index](https://mlsanity.com/paper/2609.19656/) (2026)

## Abstract (arXiv)

Existing dense retrieval models struggle with reasoning-intensive retrieval task as they fail to capture implicit relevance that requires reasoning beyond surface-level semantic information. To address these challenges, we propose Scenario-Profiled Indexing with Knowledge Expansion (SPIKE), a dense retrieval framework that explicitly indexes implicit relevance by decomposing documents into scenario-based retrieval units. SPIKE organizes documents into scenario, which encapsulates the reasoning process necessary to uncover implicit relationships between hypothetical information needs and document content. SPIKE constructs a scenario-augmented dataset using a powerful teacher large language model (LLM), then distills these reasoning capabilities into a smaller, efficient scenario generator. During inference, SPIKE incorporates scenario-level relevance alongside document-level relevance, enabling reasoning-aware retrieval. Extensive experiments demonstrate that SPIKE consistently enhances retrieval performance across various query types and dense retrievers. It also enhances the retrieval experience for users through scenario and offers valuable contextual information for LLMs in retrieval-augmented generation (RAG).