citation verification

The proliferation of misinformation on social media necessitates scalable fact-checking technologies, for which claim detection—identifying objectively verifiable statements—is a critical prerequisite. Existing approaches rely primarily on linguistic cues or claim-worthiness estimation, exhibiting limited robustness when handling ambiguous political statements and short, heterogeneous texts (e.g., tweets). To address semantic ambiguity and format heterogeneity, we propose a retrieval-augmented claim detection framework that jointly models evidence relevance structure signals and source credibility. Our method integrates external knowledge via retrieval while explicitly encoding structural relationships among candidate claims and supporting evidence, alongside calibrated source reliability scores. Evaluated on CT22-test and PoliClaim-test benchmarks, our approach significantly outperforms pure-text baselines and conventional retrieval-based methods, achieving state-of-the-art accuracy and F1-score. Results demonstrate that synergistic multi-signal modeling substantially enhances claim detection performance in complex, real-world contexts.

Existing fact-checking retrieval models primarily rely on relevance ranking, neglecting the actual discriminative utility of retrieved evidence for verifiers. Method: We propose a “utility-driven” evidence retrieval paradigm that directly enhances the support of retrieved evidence for claim verification. To this end, we design a Feedback-based Evidence Retriever (FER), which employs the KL divergence between the verifier’s (e.g., BERT-based) prediction distributions over retrieved versus gold evidence as a differentiable feedback signal, enabling end-to-end joint optimization of retrieval and verification. Our approach integrates the retriever, verifier, and KL-based feedback mechanism without requiring human-annotated evidence. Contribution/Results: On benchmarks such as FEVER, FER significantly outperforms prevailing relevance-driven baselines. It is the first work to systematically demonstrate that utility-oriented retrieval yields substantial improvements in fact verification performance, offering a novel, interpretable, and efficient pathway for automated fact-checking.

Existing evidence evaluation methods for automated fact-checking (AFC) suffer from two key limitations: they either indirectly infer evidence quality from final verdicts or rely on exact matching against closed knowledge sources (e.g., Wikipedia), resulting in narrow, low-generalizability assessments. To address this, we propose Ev2R—the first evaluation framework explicitly designed to assess evidence retrieval quality—establishing a tripartite paradigm: reference-based, proxy-reference-based, and reference-free evaluation. We innovatively introduce an LLM-driven, prompt-based scorer, rigorously validated through human evaluation alignment and adversarial testing. Experiments across multiple datasets demonstrate that our scorer achieves over 35% higher correlation with human judgments than conventional metrics (e.g., ROUGE, BERTScore) and exhibits superior robustness against noise and perturbations.

Existing AI-generated content (AIGC) fact-checking tools predominantly rely on black-box binary classification or regression models, suffering from poor interpretability, limited evidence diversity, and minimal user interactivity. Method: We propose the first user-driven, fine-grained fact verification framework that decomposes long texts into atomic claims, integrates heterogeneous multi-source evidence (e.g., knowledge bases, web pages, documents), and employs cross-source evidence fusion with an interpretable reasoning model to produce claim-level confidence scores and natural-language explanations—supporting multi-hop provenance tracing and dynamic user feedback. Contribution/Results: Our framework breaks from conventional paradigms by enabling transparent, traceable, evidence-diverse, and human-AI collaborative verification. Experiments demonstrate significant improvements in user verification efficiency (+37%) and trust (+42%), establishing a novel paradigm for trustworthy AIGC interaction.

This work investigates large language models’ (LLMs) capabilities in evidence-based claim verification, specifically evaluating deductive versus abductive reasoning. To this end, we introduce RECV—the first benchmark featuring real-world claims with fine-grained, atomic-level annotations of reasoning types—and propose a reasoning-type decomposition evaluation framework. Through systematic assessment of mainstream closed-source LLMs across multiple difficulty levels and prompting strategies, complemented by semantic similarity analysis, we find that: (1) LLMs exhibit robust performance on deductive reasoning but suffer from systematic failures in abductive reasoning; (2) generated explanations achieve high semantic similarity to human-written ones—especially for deductive tasks—but rationalization does not consistently improve verification accuracy. This study provides the first empirical evidence of LLMs’ fundamental limitations in abductive reasoning, establishing a novel, trustworthy benchmark and methodology for rigorous reasoning evaluation.

This study addresses the limitations of existing verifiable claim detection methods, which rely solely on claim text and neglect contextual information, thereby constraining verification reliability. To overcome this, the work introduces context retrieval into the task for the first time and proposes an end-to-end, context-driven detection paradigm. Specifically, it retrieves relevant evidence from Wikipedia via entity recognition and leverages large language models to generate contextual summaries that support claim classification. Experiments on the CheckThat! 2022 and PoliClaim datasets demonstrate the effectiveness of the approach, revealing that its performance is influenced by domain characteristics, model architecture, and learning settings—including fine-tuning, zero-shot, and few-shot configurations. Multidimensional analysis further elucidates the mechanisms through which contextual augmentation enhances verification accuracy.

Scientific peer review often includes subjective or unverifiable claims that compromise its fairness. This work proposes the first end-to-end claim verification system tailored for peer review comments. The system extracts verifiable claims from reviews, retrieves and re-ranks relevant evidence from the manuscript, performs fact-checking via natural language inference, and presents results through an interactive visual interface. Designed with modularity in mind, it allows flexible substitution of retrieval, re-ranking, and inference components, making it adaptable for reviewers, authors, and program committees alike. The system is publicly accessible via a live demo platform and API, accompanied by tutorials, significantly enhancing the transparency and verifiability of the peer review process.

This work addresses the challenge of automatically generating fact-checking articles grounded in verifiable citations by leveraging claims, veracity labels, and supporting evidence documents. To this end, the authors propose a multi-agent collaborative pipeline that integrates dense retrieval, source-balanced evidence selection, structured content planning, and citation-aware generation. The framework innovatively incorporates a gated self-evaluation mechanism and a natural language inference (NLI)-driven citation auditing module to repair missing citations and automatically eliminate redundant or unsupported references. Experimental results demonstrate that the proposed approach significantly improves citation accuracy and source credibility in the generated articles, thereby validating the effectiveness of jointly optimizing evidence selection, structured generation, and post-hoc citation verification.

This work addresses the critical challenge of hallucinated citations in scientific texts generated by large language models, which pose a serious threat to academic integrity and are difficult to detect manually. To this end, the authors introduce the first benchmark and verification framework specifically designed for detecting fabricated references in scientific writing. The framework features a novel unified metric for evaluating citation faithfulness and evidence alignment, supported by a large-scale, cross-domain dataset rigorously validated by human annotators. Methodologically, it proposes an interpretable and scalable multi-agent verification pipeline that integrates claim extraction, evidence retrieval, passage matching, and reasoning calibration to enable end-to-end auditing of citation authenticity. Experimental results demonstrate that the proposed approach significantly outperforms existing methods in both accuracy and interpretability, effectively identifying citation errors produced by state-of-the-art large language models and offering a reliable tool for scientific publishing integrity.

This work addresses the challenge that supporting evidence in fact-checking articles is often presented in unstructured forms, limiting its utility for automated systems. The authors propose PrimeFacts, a novel method that systematically transforms such evidence into structured, context-independent atomic premises. This is achieved by identifying hyperlink anchor texts, leveraging large language models to rewrite sentences so as to eliminate contextual dependencies, and extracting implicit evidence. The approach substantially enhances both evidence retrievability and verification performance: it yields a 30% relative improvement in Mean Reciprocal Rank for cross-article retrieval and boosts Macro-F1 scores by 10–20 percentage points on claim verification tasks. These gains are consistent across varying classification granularities and model architectures, while faithfully preserving the original source information.