Using three support-annotated multi-hop QA benchmarks, this work compares matched adapted readers trained with raw context, retrieval windows, and gold-support diagnostic renderings to distinguish support-availability failures from remaining reader-interface effects.
Abstract
In multi-hop RAG evaluation, a top-k answer score can hide two different failures: the retrieval window may drop part of the support chain, or it may contain support in a form the adapted reader does not use well. We call this reader-facing form of retrieved evidence an evidence interface. Using three support-annotated multi-hop QA benchmarks, we compare matched adapted readers trained with raw context, retrieval windows, and gold-support diagnostic renderings. These comparisons distinguish support-availability failures from remaining reader-interface effects. Top-k windows become interpretable only after checking whether the complete annotated support chain survives: when it does, short ranked windows can match or improve over raw context; when it does not, missing support explains much of the loss. Gold support-first improves matched readers; on 2Wiki and MuSiQue, a support-supervised ranker raises coverage and recovers raw-context quality at lower prompt cost, while retaining gold headroom. Support-removal checks further show that the gains rely on exposed evidence, not only answer priors. On support-annotated evaluations, top-k answer scores should therefore be reported together with complete-support coverage.
Retrieval-augmented generation (RAG) improves large language models by grounding generation in external evidence, but it also introduces a source trust problem: retrieved context may be useful, irrelevant, or even misleading. Existing RAG systems often apply a fixed trust policy toward retrieved evidence, which can either over-trust incorrect context or underuse context when the user explicitly asks for context-following behavior. Therefore, we propose Intent-Guided Decoding (IGD), a framework that arbitrates between retrieved context and parametric memory according to user intent. IGD uses answer-level filtering and token-level correction to steer the final decoding trajectory between retrieved context and parametric memory. We evaluate IGD on three faithful QA benchmarks and three factual-conflict benchmarks across five LLMs, IGD substantially improves factual recovery, achieving gains of up to 65.4 percentage points on factual-conflict benchmarks over Direct RAG, while preserving or improving strict context-following behavior, this findings highlight the importance of balancing factuality and faithfulness in RAG.
Retrieval evaluation for retrieval-augmented generation (RAG) is increasingly designed around whether retrieved passages contain evidence that can support generation, rather than topical relevance alone. We study whether this closer alignment with downstream evidence needs also makes retrieval evaluation more useful for the decisions built from it. Across five retrieval benchmarks and an end-to-end TREC RAG 2025 setting, we examine an answer-support signal in four roles: comparing retrievers, guiding retrieval training and system selection, predicting downstream answer quality, and filtering the evidence supplied to a generator. The signal changes retrieval rankings, but its downstream value is not uniform. It does not reliably improve retriever training; the benefit of using it for system selection depends on how the generator is instructed to use the retrieved evidence; and retrieval scores based on it do not robustly predict answer quality on unseen topics. In a direct evidence intervention, human annotators confirm that filtering preferentially preserves passages containing useful answer evidence, yet different answer evaluators reach different conclusions about whether the resulting answers improve. These results show that making retrieval evaluation more closely reflect the evidence needed for generation does not by itself make every downstream use of that evaluation more reliable. RAG evaluation methods should therefore be assessed with respect to the particular comparisons, decisions, and conclusions they are intended to support.
Utshab Kumar Ghosh, D. Mukhopadhyay, Shubham Chatterjee· 0 citations
Virtual Humans enhanced with Large Language Models can hold broad conversations, but their answers may sound convincing while still being factually incorrect. Such hallucinations can mislead users and reduce trust, especially because research shows that people often overestimate LLM accuracy and may remain distrustful after errors occur. This loss of trust is particularly concerning in medical contexts, where reliable information is essential. Retrieval-augmented generation (RAG) addresses this issue by grounding LLM responses in external knowledge sources that were not part of the model's training data. The described Virtual Human demonstration system uses RAG with 14 geriatrics patient brochures from the Canisius Wilhelmina Hospital, which were converted into question-answer chunks and embedded. At runtime they were retrieved based on similarity to user queries. The system allows for comparison of four RAG approaches, i.e. Basic RAG, HyDE RAG, Reranking RAG, and Cognitive RAG, each using different strategies to improve retrieval and response quality. Users can choose among these RAG options in the application menu, as well as set some other user preferences. Response quality can be evaluated using RAGAS metrics and latency related data. User experience data can also be collected for correlation to RAG performance data.
Roel Boumans· Proceedings of the 26th ACM...· 0 citations
MCoRe, a multi-entry complementary retrieval framework with reflection-guided iteration for multi-hop QA that enables multi-entry complementary retrieval by indexing entry units at multiple semantic resolutions with explicit links to chunk evidence, and fusing cross-resolution hits via chunk-level voting to form a compact evidence set for answer generation.
Juxiang Zeng, Zhuohui Gao, Zhe Hou et al.· Proceedings of the 32nd ACM...· 0 citations
Real-world image search queries are multimodal and compositional: ``find this shirt in pink''specifies an entity to retain, an attribute to modify, and context to ignore. Yet existing re-rankers either compress such multifaceted relevance into an opaque embedding or rely on free-form chain-of-thought that easily omits or hallucinates fine-grained constraints. Drawing on rubric- and checklist-based evaluation from NLP, we recast multimodal image re-ranking as a semantic constraint satisfaction problem and propose EviRank, which parses any query - text-only, image-only, or composed - into a unified evidence package: typed criteria across six semantic slots (e.g., entities, attributes, relations), each labelled required, forbidden, or ignorable. Re-ranking then reduces to evidence-conditioned verification, combining deterministic rubric scoring and evidence-grounded listwise comparison in a single training-free procedure. The explicit evidence can further serve as structured supervision for optionally distilling a lightweight student. Across five benchmarks spanning text-to-image, image-to-image, and composed image retrieval, EviRank achieves state-of-the-art performance, and the distilled student preserves over 90% of the teacher's capability at substantially lower cost.
Enjun Du, Siyi Liu, Zi-Rong Chen et al.· 1 citation
Retrieval over visually rich documents has a representation problem: important content often lives in tables, charts, figures, and layout relations that plain OCR linearizes, corrupts, or omits. ColPali-family visual retrievers address this with patch-level multi-vector indexes and late-interaction scoring, keeping image-derived retrieval on the query-time serving path. We introduce MIDR (Multimodal Indexing for Document Retrieval), a training-free framework for enrichment-augmented indexing that shifts multimodal reasoning to index time. During ingestion, a multimodal LLM converts rendered pages into verified textual fields that are indexed with BM25F and optionally fused with dense retrieval, enabling text-centric serving over multimodally grounded evidence. On ViDoRe V3, MIDR Hybrid achieves 0.6219 average nDCG across five English domains, a 23.0% relative gain over BM25, remaining competitive with ColQwen2.5. On two French-document domains, enrichment bridges English queries and French page text, lifting BM25 from 0.1532 to 0.5448 nDCG and outperforming ColQwen2.5. Across all seven domains, MIDR leads ColQwen2.5 on four while using approximately 9x smaller index memory and approximately 2x lower query latency. These results establish index-time multimodal reasoning as a compelling accuracy-deployment alternative to serving-time visual late interaction.
Debanjan Mahata, Atharva Tendle, Daniel Preoţiuc-Pietro et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
