# Verifiable Generation with Subsentence-Level Fine-Grained Citations

arXiv: https://arxiv.org/abs/2406.06125 · HTML: https://mlsanity.com/paper/2406.06125/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 15 references that are in the map.
- First posted on arXiv 10 Jun 2024.

## Builds on (most cited first, 11 of 11)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation](https://mlsanity.com/paper/2305.14251/) (2023)
- [Enabling Large Language Models to Generate Text with Citations](https://mlsanity.com/paper/2305.14627/) (2023)
- [Evaluating Verifiability in Generative Search Engines](https://mlsanity.com/paper/2304.09848/) (2023)
- [Teaching language models to support answers with verified quotes](https://mlsanity.com/paper/2203.11147/) (2022)
- [Measuring Attribution in Natural Language Generation Models](https://mlsanity.com/paper/2112.12870/) (2021)
- [Revisiting the Gold Standard: Grounding Summarization Evaluation with Robust Human Evaluation](https://mlsanity.com/paper/2212.07981/) (2022)
- [WiCE: Real-World Entailment for Claims in Wikipedia](https://mlsanity.com/paper/2303.01432/) (2023)
- [LLatrieval: LLM-Verified Retrieval for Verifiable Generation](https://mlsanity.com/paper/2311.07838/) (2023)
- [Source-Aware Training Enables Knowledge Attribution in Language Models](https://mlsanity.com/paper/2404.01019/) (2024)

## Built on by (most cited first, 15 of 15)

- [From Lossy to Verified: A Provenance-Aware Tiered Memory for Agents](https://mlsanity.com/paper/2602.17913/) (2026)
- [Credible Plan-Driven RAG Method for Multi-Hop Question Answering](https://mlsanity.com/paper/2504.16787/) (2025)
- [TROVE: A Challenge for Fine-Grained Text Provenance via Source Sentence Tracing and Relationship Classification](https://mlsanity.com/paper/2503.15289/) (2025)
- [GenProve: Learning to Generate Text with Fine-Grained Provenance](https://mlsanity.com/paper/2601.04932/) (2026)
- [Are Finer Citations Always Better? Rethinking Granularity for Attributed Generation](https://mlsanity.com/paper/2604.01432/) (2026)
- [Legal LLM Hallucination Should Be Evaluated as Failure of Legal Warrant](https://mlsanity.com/paper/2609.17546/) (2026)
- [Rate, Explain and Cite (REC): Enhanced Explanation and Attribution in Automatic Evaluation by Large Language Models](https://mlsanity.com/paper/2411.02448/) (2024)
- [RUBEN: Rule-Based Explanations for Retrieval-Augmented LLM Systems](https://mlsanity.com/paper/2605.10862/) (2026)
- [Issue Retrieval and Verification Enhanced Supplementary Code Comment Generation](https://mlsanity.com/paper/2506.14649/) (2025)
- [LCDS: A Logic-Controlled Discharge Summary Generation System Supporting Source Attribution and Expert Review](https://mlsanity.com/paper/2507.05319/) (2025)
- [Rule-Based Explanations for Retrieval-Augmented LLM Systems](https://mlsanity.com/paper/2510.22689/) (2025)
- [CC30k: A Citation Contexts Dataset for Reproducibility-Oriented Sentiment Analysis](https://mlsanity.com/paper/2511.07790/) (2025)
- [Plans for Evaluating Structured Generative Search Summaries](https://mlsanity.com/paper/2605.26400/) (2026)
- [Explicit Evidence Grounding via Structured Inline Citation Generation](https://mlsanity.com/paper/2606.07130/) (2026)
- [Tokengeist: Multi-Turn Attribution Tracing in Agentic Conversations](https://mlsanity.com/paper/2607.22610/) (2026)

## Abstract (arXiv)

Verifiable generation requires large language models (LLMs) to cite source documents supporting their outputs, thereby improve output transparency and trustworthiness. Yet, previous work mainly targets the generation of sentence-level citations, lacking specificity about which parts of a sentence are backed by the cited sources. This work studies verifiable generation with subsentence-level fine-grained citations for more precise location of generated content supported by the cited sources. We first present a dataset, SCiFi, comprising 10K Wikipedia paragraphs with subsentence-level citations. Each paragraph is paired with a set of candidate source documents for citation and a query that triggers the generation of the paragraph content. On SCiFi, we evaluate the performance of state-of-the-art LLMs and strategies for processing long documents designed for these models. Our experiment results reveals key factors that could enhance the quality of citations, including the expansion of the source documents' context accessible to the models and the implementation of specialized model tuning.