Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents

• URL: http://arxiv.org/abs/2604.03173v1
• Date: Fri, 03 Apr 2026 16:49:02 GMT
• Title: Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents
• Abstract summary: Large language models and deep research agents supply citation URLs to support their claims.<n>We address six research questions about citation URL validity using 10 models and agents on DRBench and 3 models on ExpertQA.<n>We find that 3--13% of citation URLs are hallucinated, while 5--18% are non-resolving overall.<n>We release URLhealth, an open-source tool for URL liveness checking and stale-vs-hallucinated classification using the Wayback Machine.
• Score: 38.68399248551709
• License: http://creativecommons.org/publicdomain/zero/1.0/
• Abstract: Large language models and deep research agents supply citation URLs to support their claims, yet the reliability of these citations has not been systematically measured. We address six research questions about citation URL validity using 10 models and agents on DRBench (53,090 URLs) and 3 models on ExpertQA (168,021 URLs across 32 academic fields). We find that 3--13\% of citation URLs are hallucinated -- they have no record in the Wayback Machine and likely never existed -- while 5--18\% are non-resolving overall. Deep research agents generate substantially more citations per query than search-augmented LLMs but hallucinate URLs at higher rates. Domain effects are pronounced: non-resolving rates range from 5.4\% (Business) to 11.4\% (Theology), with per-model effects even larger. Decomposing failures reveals that some models fabricate every non-resolving URL, while others show substantial link-rot fractions indicating genuine retrieval. As a solution, we release urlhealth, an open-source tool for URL liveness checking and stale-vs-hallucinated classification using the Wayback Machine. In agentic self-correction experiments, models equipped with urlhealth reduce non-resolving citation URLs by $6\textrm{--}79\times$ to under 1\%, though effectiveness depends on the model's tool-use competence. The tool and all data are publicly available. Our characterization findings, failure taxonomy, and open-source tooling establish that citation URL validity is both measurable at scale and correctable in practice.

Related papers

• Who is the Agent to Blame? Localizing Faithfulness and Citation Mistakes in Agentic Deep Research [75.8770212601511]
  Deep research (DR) systems produce long-form cited reports by orchestrating multiple agents that search and synthesize information from the web.<n> Citations are the primary mechanism for evaluating the faithfulness of these reports, yet current DR systems exhibit poor citation recall.<n>We propose an evaluation method that pinpoints which agent introduced each error by locally testing agent invocations for faithfulness and verifiability relative to their own inputs.
  arXiv  Detail & Related papers  (2026-08-25T09:34:18Z)
• Verified Misguidance: Measuring Structural Citation Failures in Search-Augmented LLMs [8.721564756242431]
  CITETRACE is a dataset that traces the full citation chain from user query through retrieved source to generated answer.<n>We design a three-dimension evaluation framework that scores each citation on intent-purpose alignment, source suitability, and answer-source fidelity.<n>Across our pool, 30.6% of citations distort their sources and 27.1% originate from domain-inappropriate sources.
  arXiv  Detail & Related papers  (2026-05-27T14:54:05Z)
• Source or It Didn't Happen: A Multi-Agent Framework for Citation Hallucination Detection [30.352741291341843]
  Existing detectors reduce to binary found/not-found decisions, offering little field-level signal to auditors.<n>We introduce a 12-code taxonomy spanning Real, Potential, and Hallucinated citations.<n>CiteTracer extracts citations from PDF and Bib, fetches evidence through lookup URL, scholar connectors, and web search.<n>It reaches 97.1% accuracy on the synthetic benchmark, with class-level F1 scores of 97.0, 95.8, and 98.5 for Real, Potential, and Hallucinated, respectively, and fabricates hallucinations on the real-world set without abstaining.
  arXiv  Detail & Related papers  (2026-05-09T00:53:24Z)
• BibTeX Citation Hallucinations in Scientific Publishing Agents: Evaluation and Mitigation [34.429649156970015]
  Prior evaluations tested base models without search, which does not reflect current practice.<n>Three search-enabled frontier models generate Bib entries scored on nine fields and a six-way error taxonomy.<n>Overall accuracy is 83.6%, but only 50.9% of entries are fully correct.
  arXiv  Detail & Related papers  (2026-04-03T16:30:58Z)
• CiteAudit: You Cited It, But Did You Read It? A Benchmark for Verifying Scientific References in the LLM Era [51.63024682584688]
  Large language models (LLMs) introduce a new risk: fabricated references that appear plausible but correspond to no real publications.<n>We present the first comprehensive benchmark and detection framework for hallucinated citations in scientific writing.<n>Our framework significantly outperforms prior methods in both accuracy and interpretability.
  arXiv  Detail & Related papers  (2026-02-26T19:17:39Z)
• How LLMs Cite and Why It Matters: A Cross-Model Audit of Reference Fabrication in AI-Assisted Academic Writing and Methods to Detect Phantom Citations [1.0829694003408499]
  Large language models (LLMs) have been noted to fabricate scholarly citations, yet the scope of this behavior remains poorly quantified.<n>We present one of the largest hallucination citation audits to date, in which 10 commercially deployed LLMs were prompted across four academic domains.<n>Our results show that the observed hallucination rates span a fivefold range (between 11.4% and 56.8%) and are strongly shaped by model, domain, and prompt framing.
  arXiv  Detail & Related papers  (2026-02-07T00:14:00Z)
• GhostCite: A Large-Scale Analysis of Citation Validity in the Age of Large Language Models [22.147294042024836]
  Citations provide the basis for trusting scientific claims; when they are invalid or fabricated, this trust collapses.<n>With the advent of Large Language Models (LLMs), this risk has intensified.<n>We develop CiteVerifier, an open-source framework for large-scale citation verification.
  arXiv  Detail & Related papers  (2026-02-06T14:08:34Z)
• The 17% Gap: Quantifying Epistemic Decay in AI-Assisted Survey Papers [0.0]
  "Hallucinated papers" are a known artifact, but the systematic degradation of valid citation chains remains unquantified.<n>We conducted a forensic audit of 50 recent survey papers in Artificial Intelligence published between September 2024 and January 2026.<n>We detect a persistent 17.0% Phantom Rate -- citations that cannot be resolved to any digital object despite aggressive forensic recovery.
  arXiv  Detail & Related papers  (2026-01-24T12:00:55Z)
• Citation Failure: Definition, Analysis and Efficient Mitigation [56.09968229868067]
  Citations from LLM-based RAG systems are supposed to simplify response verification.<n>This does not hold for citation failure, when a model generates a helpful response, but fails to cite complete evidence.<n>We propose to disentangle this from response failure, where the response itself is flawed, and citing complete evidence is impossible.
  arXiv  Detail & Related papers  (2025-10-23T07:47:22Z)
• WebCiteS: Attributed Query-Focused Summarization on Chinese Web Search Results with Citations [34.99831757956635]
  We formulate the task of attributed query-focused summarization (AQFS) and present WebCiteS, a Chinese dataset featuring 7k human-annotated summaries with citations. We tackle these issues by developing detailed metrics and enabling the automatic evaluator to decompose the sentences into sub-claims for fine-grained verification.
  arXiv  Detail & Related papers  (2024-03-04T07:06:41Z)
• Evaluating Verifiability in Generative Search Engines [70.59477647085387]
  Generative search engines directly generate responses to user queries, along with in-line citations. We conduct human evaluation to audit four popular generative search engines. We find that responses from existing generative search engines are fluent and appear informative, but frequently contain unsupported statements and inaccurate citations.
  arXiv  Detail & Related papers  (2023-04-19T17:56:12Z)
• Deep Graph Learning for Anomalous Citation Detection [55.81334139806342]
  We propose a novel deep graph learning model, namely GLAD (Graph Learning for Anomaly Detection), to identify anomalies in citation networks. Within the GLAD framework, we propose an algorithm called CPU (Citation PUrpose) to discover the purpose of citation based on citation texts.
  arXiv  Detail & Related papers  (2022-02-23T09:05:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.