# MultAttnAttrib: Training-Free Multimodal Attribution in Long Document Question Answering

arXiv: https://arxiv.org/abs/2607.01420 · HTML: https://mlsanity.com/paper/2607.01420/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 134 references that are in the map.
- First posted on arXiv 1 Jul 2026.

## Builds on (most cited first, 20 of 20)

- [Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection](https://mlsanity.com/paper/2310.11511/) (2023)
- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [What Does BERT Look At? An Analysis of BERT's Attention](https://mlsanity.com/paper/1906.04341/) (2019)
- [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)
- [Retrieval Head Mechanistically Explains Long-Context Factuality](https://mlsanity.com/paper/2404.15574/) (2024)
- [Stabilizing Transformer Training by Preventing Attention Entropy Collapse](https://mlsanity.com/paper/2303.06296/) (2023)
- [Attribute First, then Generate: Locally-attributable Grounded Text Generation](https://mlsanity.com/paper/2403.17104/) (2024)
- [MINT-1T: Scaling Open-Source Multimodal Data by 10x: A Multimodal Dataset with One Trillion Tokens](https://mlsanity.com/paper/2406.11271/) (2024)
- [Training Language Models to Generate Text with Citations via Fine-grained Rewards](https://mlsanity.com/paper/2402.04315/) (2024)
- [Benchmarking Retrieval-Augmented Multimodal Generation for Document Question Answering](https://mlsanity.com/paper/2505.16470/) (2025)
- [VISA: Retrieval Augmented Generation with Visual Source Attribution](https://mlsanity.com/paper/2412.14457/) (2024)
- [Peering into the Mind of Language Models: An Approach for Attribution in Contextual Question Answering](https://mlsanity.com/paper/2405.17980/) (2024)
- [Beyond Logit Lens: Contextual Embeddings for Robust Hallucination Detection & Grounding in VLMs](https://mlsanity.com/paper/2411.19187/) (2024)
- [Learning to Generate Answers with Citations via Factual Consistency Models](https://mlsanity.com/paper/2406.13124/) (2024)
- [CoTAR: Chain-of-Thought Attribution Reasoning with Multi-level Granularity](https://mlsanity.com/paper/2404.10513/) (2024)
- [Enhancing Post-Hoc Attributions in Long Document Comprehension via Coarse Grained Answer Decomposition](https://mlsanity.com/paper/2409.17073/) (2024)
- [VeriCite: Towards Reliable Citations in Retrieval-Augmented Generation via Rigorous Verification](https://mlsanity.com/paper/2510.11394/) (2025)
- [SciClaimEval: Cross-modal Claim Verification in Scientific Papers](https://mlsanity.com/paper/2602.07621/) (2026)
- [Multimodal Fact-Level Attribution for Verifiable Reasoning](https://mlsanity.com/paper/2602.11509/) (2026)

## Abstract (arXiv)

As grounded QA systems are increasingly deployed in AI assistants, accurately attributing generated answers to evidence is critical for user trust and model safety. While unimodal attributions have been explored in depth, the multimodal setting remains relatively under-researched. As a result, we introduce MultAttnAttrib, a training-free attribution-generation method that leverages a model's prefill pass, selected attention heads, and calibrated thresholds to locate source evidence within a document. To establish baseline results for the method, we introduce MultAttrEval, a complementary benchmark dataset annotated with fine-grained, ground-truth attributions for answer components grounded in multimodal source documents. To our knowledge, this is the first evaluation dataset designed specifically for multimodal attribution in long-form documents. Experimental results show that MultAttnAttrib consistently outperforms a variety of attribution-generation methods, including several strong prompting-based approaches and matches the latest frontier models such as GPT 5.4. Our method not only substantially improves attribution accuracy for both unimodal and multimodal attribution types, but also produces attributions at up to one-seventh of the direct inference latency compared to prompting on the same base model.