# Retrieval-in-the-Chain: Bootstrapping Large Language Models for Generative Retrieval

arXiv: https://arxiv.org/abs/2510.13095 · HTML: https://mlsanity.com/paper/2510.13095/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning.
- Builds on 19 of its 35 references that are in the map.
- First posted on arXiv 15 Oct 2025.

## Placed in

- LLM reasoning / [Knowledge- and tool-augmented reasoning](https://mlsanity.com/lr/knowledge-and-tool-augmented-reasoning/)
- Synthetic data and self-improvement for LLMs / [LLM synthetic data generation](https://mlsanity.com/synth/llm-synthetic-data-generation/)

## Builds on (most cited first, 19 of 19)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Self-Consistency Improves Chain of Thought Reasoning in Language Models](https://mlsanity.com/paper/2203.11171/) (2022)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [MS MARCO: A Human Generated MAchine Reading COmprehension Dataset](https://mlsanity.com/paper/1611.09268/) (2016)
- [ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT](https://mlsanity.com/paper/2004.12832/) (2020)
- [Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval](https://mlsanity.com/paper/2007.00808/) (2020)
- [Transformer Memory as a Differentiable Search Index](https://mlsanity.com/paper/2202.06991/) (2022)
- [Document Expansion by Query Prediction](https://mlsanity.com/paper/1904.08375/) (2019)
- [Qwen2.5-1M Technical Report](https://mlsanity.com/paper/2501.15383/) (2025)
- [Autoregressive Search Engines: Generating Substrings as Document Identifiers](https://mlsanity.com/paper/2204.10628/) (2022)
- [Learning to Rank in Generative Retrieval](https://mlsanity.com/paper/2306.15222/) (2023)
- [Scalable and Effective Generative Information Retrieval](https://mlsanity.com/paper/2311.09134/) (2023)
- [CorpusBrain: Pre-train a Generative Retrieval Model for Knowledge-Intensive Language Tasks](https://mlsanity.com/paper/2208.07652/) (2022)
- [Planning Ahead in Generative Retrieval: Guiding Autoregressive Generation through Simultaneous Decoding](https://mlsanity.com/paper/2404.14600/) (2024)
- [Generative Retrieval via Term Set Generation](https://mlsanity.com/paper/2305.13859/) (2023)

## Built on by (most cited first, 1 of 1)

- [Integrating Chain-of-Thought into Generative Retrieval: A Preliminary Study](https://mlsanity.com/paper/2605.22358/) (2026)

## Abstract (arXiv)

Generative retrieval (GR) is an emerging paradigm that leverages large language models (LLMs) to autoregressively generate document identifiers (docids) relevant to a given query. Prior works have focused on leveraging the generative capabilities of LLMs to improve GR, while overlooking that their reasoning capabilities could likewise help. This raises a key question: Can explicit reasoning benefit GR? To investigate, we first conduct a preliminary study where an LLM is prompted to generate free-form chain-of-thought (CoT) reasoning before performing constrained docid decoding. Although this method outperforms standard GR, the generated reasoning tends to be verbose and poorly aligned with the docid space. These limitations motivate the development of a reasoning mechanism better tailored to GR. Therefore, we propose Reason-for-Retrieval (R4R), a reasoning-augmented framework for GR that converts free-form CoT reasoning into a compact, structured format, and iteratively refines the reasoning during the retrieval process. R4R augments an existing GR method by leveraging a reasoning-capable LLM that has been instruction-tuned for GR. At inference time, R4R first uses the LLM to generate an initial structured reasoning; then the same LLM alternates between (i) constrained decoding with the chosen GR method to produce candidate docids and (ii) updating the reasoning based on retrieval results to improve the next round. R4R does not require additional models or training, and instead a single LLM serves as both the reasoning generator and the retriever. Extensive experiments on Natural Questions, MS MARCO, and a real-world item-search benchmark validate the effectiveness of R4R.