Pith. sign in

REVIEW 3 major objections 6 minor 50 references

Fact in Fragments: Deconstructing Complex Claims via LLM-based Atomic Fact Extraction and Verification

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read By iteratively decomposing complex claims into atomic facts and feeding each verified label and rationale back into the next extraction step, AFEV claims state-of-the-art accuracy and interpretability on five fact-verification datasets.

desk verdict A solid, well-written pipeline paper with small but consistent accuracy gains, whose "interpretability" claim is never actually measured. read the letter →

arxiv 2506.07446 v1 pith:IACU2NAK submitted 2025-06-09 cs.AI

Liwen Zheng , Chaozhuo Li , Zheng Liu , Feiran Huang , Haoran Jia , Zaisheng Ye , Xi Zhang This is my paper
classification cs.AI
keywords factverificationatomicextractionmulti-hopreasoningevidenceretrievallargelanguagemodelsiterativedecompositioninterpretabilityclaim
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that the hard part of verifying complex claims is decomposition, and that decomposition should be iterative and driven by verification feedback rather than performed once. It claims AFEV, a pipeline that extracts one atomic fact at a time, retrieves and reranks evidence for that fact, verifies it with an LLM reasoner, and feeds the label and rationale back into the next extraction step, outperforms both multi-granular supervised systems and LLM reasoning baselines. Across LIAR-PLUS, HOVER, PolitiHop, RAWFC, and LIAR, the reported gains are state-of-the-art on label accuracy, macro-F1, and retrieval F1. A sympathetic reader should care because the method promises to turn unverifiable long claims into a sequence of small, inspectable verification steps with a rationale for each step.

What carries the argument

The load-bearing mechanism is the iterative extraction loop defined in Eq. (2): $F_t = \text{Extractor}(C, F_{1:t-1}, y_{1:t-1}, r_{1:t-1})$, where $F_t$ is the next atomic fact, $C$ is the original claim, and $y_{1:t-1}, r_{1:t-1}$ are the labels and rationales from earlier verifications. Around this loop the pipeline places a retrieve-then-rerank step that keeps only the top evidence pieces for each atomic fact, a dynamic instance retriever that fetches the most similar training claim as an in-context demonstration, and a reasoner that returns both a label and a rationale. Atomic facts are compact, independently verifiable units, which is what allows evidence and demonstrations to stay small and precise. The loop terminates once the extracted facts cover the claim, so the number of iterations is controlled.

What would settle it

Take a set of complex claims, run AFEV once normally, and run it again after manually replacing one entity in the first rationale with a plausible but false entity before the second extraction step. If final verification accuracy stays the same, Eq. (2)'s feedback is not doing the load-bearing work the paper attributes to it; if accuracy changes, the loop propagates rationale errors. Either result settles whether the feedback loop is the source of the reported gains.

Watch

Extended reading notes

Core claim

The central discovery is that iterative atomic fact extraction with feedback reduces error accumulation in complex claim verification. Instead of decomposing a claim once and then retrieving evidence for all fragments at once, the extractor generates the next atomic fact conditioned on the original claim, the previously extracted facts, and their verification labels and rationales. The reasoner simultaneously produces a factuality label and an evidence-grounded rationale per atomic fact; the rationale can introduce implicit information, such as an entity named only in earlier evidence, that sharpens the next decomposition. The paper reports that this feedback loop, together with an evidence reranker and dynamically retrieved demonstrations, achieves state-of-the-art results on all five datasets, including 83.73 label accuracy on LIAR-PLUS, 78.87 on HOVER, and 74.14 on PolitiHop, and 60.2 F1 on RAWFC and 43.9 F1 on LIAR.

Load-bearing premise

The feedback loop assumes the labels and rationales generated for earlier atomic facts are reliable enough to condition later extraction; if an earlier rationale contains a hallucinated entity or a wrong relationship, subsequent atomic facts can be extracted from that contaminated information, and errors can accumulate instead of shrinking.

Editorial extensions

If this is right

  • Complex claims can be verified as a sequence of smaller, independently checkable facts rather than as one monolithic judgment.
  • Feeding verification labels and rationales back into extraction reduces redundant or broken atomic facts, which the paper identifies as the mechanism behind the accuracy gains.
  • Reranking evidence per atomic fact means each sub-claim needs only one or two precise evidence pieces, lowering noise compared with retrieving for the whole claim.
  • The pipeline outputs a label and rationale for every atomic fact, so the final verdict can be traced back to individual evidence-grounded steps.
  • On the five benchmarks tested, the reported numbers place the method ahead of both multi-granular supervised baselines and the strongest LLM prompting baselines.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An untested extension would measure decomposition fidelity directly by checking whether atomic facts plus rationales reconstruct the original claim; the paper reports no such metric.
  • The feedback loop's safety depends on rationale quality, so a controlled corruption test of rationales would reveal whether the loop reduces errors or silently propagates hallucinated entities.
  • Because the extractor and reasoner are untrained LLM calls, swapping in a different base model could test whether the iterative structure or the particular model carries the gains.
  • The same iterative decomposition could transfer to long-form question answering or claim generation, where verifying one component should inform the next; the paper does not test these settings.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes AFEV, a fact-verification framework that iteratively decomposes a complex claim into atomic facts, conditions each new atomic fact on previously extracted facts and their labels and rationales (Eq. 2), retrieves and reranks evidence per atomic fact, dynamically selects demonstrations, and aggregates per-fact verdicts into a final label (Eq. 10). The framework is evaluated on HOVER, PolitiHop, LIAR-PLUS, RAWFC, and LIAR, with additional retrieval, ablation, hyperparameter-sensitivity, and efficiency studies. The authors claim state-of-the-art accuracy and interpretability in the abstract and conclusion.

Significance. If the reported accuracy numbers are stable, AFEV is a useful contribution to LLM-based fact verification: it combines iterative decomposition with lightweight evidence reranking and dynamic demonstrations, and it shows gains over strong baselines such as CURE, VMASK, and RAFTS on five benchmarks. The experimental design is broad, the ablations cover all major modules, and the paper gives concrete prompts and implementation details. However, the interpretability half of the headline claim is not evaluated at all, and the case study in Section 5.7 illustrates a contamination mechanism through Equation (2) that could undermine the claimed error-reduction benefit. The accuracy claim itself rests on small margins and an underspecified significance test. These issues are fixable but require additional evidence or a rescoping of the claims.

major comments (3)
  1. [§5.7, Eq. (2), Eq. (10)] The interpretability claim and the claim that iterative feedback reduces error propagation are not supported. Equation (2) conditions the extraction of the next atomic fact on previous rationales r_{1:t-1}; the case study in §5.7 explicitly shows that Fact3 includes 'Lionel Messi' because it appeared in the rationale of Fact2, not in the original claim. If a rationale contains a hallucinated entity or relation, that content enters later atomic facts and is ultimately passed to the final aggregation in Eq. (10), so the system can end up judging a statement that is not a decomposition of the original claim. The paper reports no human evaluation of rationales, no faithfulness metric, and no estimate of how often this contamination occurs. Therefore the 'interpretability' half of the headline claim, and the claim that the feedback loop reduces rather than amplifies errors, are unverified. The authors should either provide direct measurements of rationale faithfulness and contamination frequency or remove/qualify the interpretability and error-reduction claims.
  2. [§5.2, Tables 2 and 3] The statistical evidence for the state-of-the-art accuracy claim is weak as reported. The tables state that the improvement over the best baseline is statistically significant by a sign test with p<0.01, but the test is not described: there is no indication of the number of paired samples, whether the test is per claim or per dataset, or whether it accounts for multiple comparisons. The observed margins are small (e.g., LIAR F1 43.9 vs 42.0 in Table 3; PolitiHop label accuracy 74.14 vs 72.34 in Table 2), and no error bars, standard deviations, or multi-seed results are given. Because these numbers are the sole support for the central accuracy claim, the authors should provide a precise account of the significance test, report variance over independent runs, and state whether hyperparameters such as k=2 and d=1 were fixed before evaluating all datasets.
  3. [§5.4, Table 5] The ablation supporting the rationale feedback loop is overstated. Removing rationales from the extractor input changes HOVER label accuracy from 78.87 to 78.74, a decrease of 0.13 point, while removing iterative extraction entirely decreases accuracy by 1.21 points and removing atomic-fact extraction by 1.83 points. The text's statement that removal of each component leads to a 'noticeable decline' is therefore inaccurate for the rationales component. Moreover, if the feedback loop is intended to reduce error propagation, the 0.13-point effect and the contamination risk shown in §5.7 should be analyzed with error bars and an error analysis before concluding that the mechanism is beneficial.
minor comments (6)
  1. [Tables 5 and 6] The 'w/o Iterative Extraction' ablation reports label accuracy 77.66 and Macro-F1 77.81 in Table 5, but label accuracy 77.04 and Macro-F1 76.96 in Table 6. Please reconcile these numbers.
  2. [§5.3, Table 4] The sentence that AFEV achieves state-of-the-art retrieval performance across all datasets is not supported by the table, since on HOVER the Pipeline baseline has a higher retrieval F1 (0.9427 vs 0.9423). Please correct the claim or clarify the metric being compared.
  3. [Table 2 caption] The caption identifies VMASK as the best baseline, but on HOVER the best baseline is CURE (76.98 vs 74.38). Please correct the caption or the significance description.
  4. [§5.1.4 and §5.5] Please specify the exact GPT-3.5 model version, decoding temperature, and the number of initial retrieval candidates k'; also report whether k=2 and d=1 were tuned only on HOVER and then frozen for the other datasets, and provide standard deviations over multiple runs.
  5. [§4.4] The symbol d is used both for the number of dynamic demonstrations in Eq. (8) and for the embedding dimension in the complexity expression O(k'd); please disambiguate.
  6. [Title page] The header 'Preprint submitted to Nuclear Physics B' appears to be a leftover template line and should be removed or corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: held-out benchmark comparisons and component ablations do not reduce to fitted inputs or to self-citations.

full rationale

The paper's central claim is state-of-the-art accuracy and interpretability on five fact-verification benchmarks. The accuracy comparison is evaluated on held-out test sets (Tables 2-3) against baselines, and the hyperparameters (k, d) are fixed and analyzed on HOVER; there is no evidence that the reranker or reasoner was trained on test labels. The reranker is trained with LLM-generated relevance labels, but the final verification labels come from the benchmark datasets, so the end-task metric is not defined in terms of the model's own outputs. Eq. (2) conditions the extractor on prior rationales r_{1:t-1}, and Eq. (9) feeds rationales back, but this is a design mechanism, not a definitional equivalence: the extracted facts and final label are not equal to the rationales by construction, and the ablation 'w/o Rationales' shows a small but non-forced difference. The case study in Sec. 5.7 does reveal a robustness risk: Fact3 incorporates 'Lionel Messi' from the rationale of Fact2 rather than from the original claim, which could propagate hallucinated content; that is a validity/contamination concern, not a circularity. Likewise, the abstract's 'interpretability' claim is not directly measured by any faithfulness metric or human evaluation, but absence of evidence is not circularity. The only self-citations (Refs. [4] and [6]) are related-work citations and are not load-bearing for the method's derivation or its claimed results. Under the rule that circularity requires a specific reduction to the paper's own inputs or a load-bearing self-citation chain, no such step is present.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim depends on GPT-3.5 as extractor, reasoner, and reducer of training labels; on a dense retrieval model and reranker whose training data are LLM judgments; and on a handful of hand-tuned hyperparameters. No code or trained models are released. No new entities are introduced.

free parameters (4)
  • k (number of reranked evidence pieces) = 2
    Set after extensive parameter tuning on HOVER; sensitivity analysis in Fig. 5(a).
  • k' (number of initial retrieval candidates) = 5
    Reported in Section 5.1.4 as final setting after tuning; not included in sensitivity analysis.
  • d (number of dynamic demonstrations) = 1
    Set after tuning; sensitivity analysis in Fig. 5(b) shows 1-2 is best.
  • InfoNCE temperature tau = not reported
    Appears in Eq. (6) for reranker training; no value or tuning reported.
assumptions (5)
  • domain assumption Dense cosine similarity is an adequate relevance signal for evidence and demonstration retrieval.
    Used in Eqs. (3)-(4) and (7)-(8); no retrieval quality analysis beyond final accuracy.
  • domain assumption GPT-3.5 can reliably decompose claims into atomic facts, assess evidence relevance for reranker training, and produce correct veracity labels and rationales.
    GPT-3.5 serves as extractor, reasoner, and LLM judge for positive/negative sampling in Section 4.2.1; no human agreement or error analysis is provided.
  • domain assumption Retrieved training-set demonstrations with gold labels and rationales improve reasoning without leaking the test answer.
    Dynamic demonstrations include a claim, its ground-truth label, gold evidence, and a rationale (Section 3); leakage risk across the 1:1 sampled candidate pool is not analyzed.
  • domain assumption The LLM-based termination decision, whether extracted facts cover the claim, is reliable.
    Section 4.1 relies on the extractor judging coverage to stop or continue; no evaluation of termination quality is provided.
  • ad hoc to paper The candidate pool for demonstrations is built by randomly sampling training instances at a 1:1 ratio with the test set.
    Section 4.4 introduces this ratio to cut retrieval cost; it is a design choice not derived from any principle.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fact in Fragments: Deconstructing Complex Claims via LLM-based Atomic Fact Extraction and Verification." pith.science (2026). https://pith.science/paper/IACU2NAK

@misc{pith2026250607446,
  author       = {Pith},
  title        = {Pith review of: Fact in Fragments: Deconstructing Complex Claims via LLM-based Atomic Fact Extraction and Verification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IACU2NAK}},
  note         = {Machine review of arXiv:2506.07446}
}
read the original abstract

Fact verification plays a vital role in combating misinformation by assessing the veracity of claims through evidence retrieval and reasoning. However, traditional methods struggle with complex claims requiring multi-hop reasoning over fragmented evidence, as they often rely on static decomposition strategies and surface-level semantic retrieval, which fail to capture the nuanced structure and intent of the claim. This results in accumulated reasoning errors, noisy evidence contamination, and limited adaptability to diverse claims, ultimately undermining verification accuracy in complex scenarios. To address this, we propose Atomic Fact Extraction and Verification (AFEV), a novel framework that iteratively decomposes complex claims into atomic facts, enabling fine-grained retrieval and adaptive reasoning. AFEV dynamically refines claim understanding and reduces error propagation through iterative fact extraction, reranks evidence to filter noise, and leverages context-specific demonstrations to guide the reasoning process. Extensive experiments on five benchmark datasets demonstrate that AFEV achieves state-of-the-art performance in both accuracy and interpretability.

Figures

Figures reproduced from arXiv: 2506.07446 by the authors.

Figure 1
Figure 1. Comparison of different claim decomposition strategies. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Framework of the proposed AFEV model. eliminate irrelevant information and retrieve highly relevant instances as dy￾namic reasoning demonstrations; and 3) the Adaptive Atomic Fact Verifica￾tion module leverages the retrieved evidence and demonstrations to validate atomic facts, extracting key evidence to support the verification process and provide interpretable rationales for the results. 4.1. Dynamic Atomic Fact E… view at source ↗
Figure 3
Figure 3. Prompt for dynamic atomic fact extraction. [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Prompt for adaptive fact verification. evidence. These methods not only aggregate a richer set of clues by leveraging both sentence-level and token-level rationales, but also reduce the reasoning burden on the model through the explicit use of token-level evidence, whi…
Figure 5
Figure 5. Figure 5: Hyperparameter sensitivity analysis. results in insufficient contextual guidance, while too many (e.g., d > 2) intro￾duce redundancy and unnecessary complexity. Results in [PITH_FULL_IMAGE:figures/full_fig_p023_5.png]
Figure 6
Figure 6. Figure 6: The reasoning process of AFEV for a specific case. [PITH_FULL_IMAGE:figures/full_fig_p024_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

50 extracted references · 41 canonical work pages

  1. [1]

    Thorne, A

    J. Thorne, A. Vlachos, C. Christodoulopoulos, A. Mittal, FEVER: a large-scale dataset for fact extraction and VERification, in: Proceedings of the 2018 Conference of the North American Chapter of the Associ- ation for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), Association for Computational Linguistics, New Orleans, Lou...

    work page 2018

  2. [2]

    Z. Guo, M. Schlichtkrull, A. Vlachos, A survey on automated fact- checking, Transactions of the Association for Computational Linguistics 10 (2022) 178–206

    work page 2022

  3. [3]

    Ostrowski, A

    W. Ostrowski, A. Arora, P. Atanasova, I. Augenstein, Multi-hop fact checking of political claims, in: Proceedings of the Thirtieth Interna- tional Joint Conference on Artificial Intelligence, IJCAI-21, Interna- tional Joint Conferences on Artificial Intelligence Organization, 2021, pp. 3892–3898, main Track

    work page 2021

  4. [4]

    Zheng, C

    L. Zheng, C. Li, X. Zhang, Y.-M. Shang, F. Huang, H. Jia, Evidence retrieval is almost all you need for fact verification, in: Findings of the Association for Computational Linguistics ACL 2024, 2024, pp. 9274– 9281

    work page 2024

  5. [5]

    Khattab, C

    O. Khattab, C. Potts, M. A. Zaharia, Baleen: Robust multi-hop rea- soning at scale via condensed retrieval, in: Advances in Neural Informa- tion Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, NeurIPS 2021, December 6-14, 2021, virtual, 2021, pp. 27670–27682

    work page 2021

  6. [6]

    Zheng, C

    L. Zheng, C. Li, L. Zhang, H. Jia, S. Wang, Z. Liu, X. Zhang, Mrr- fv: Unlocking complex fact verification with multi-hop retrieval and reasoning, Proceedings of the AAAI Conference on Artificial Intelligence 39 (2025) 26066–26074. doi:10.1609/aaai.v39i24.34802. 26

    work page doi:10.1609/aaai.v39i24.34802 2025

  7. [7]

    H. Li, X. Han, H. Wang, Y. Wang, M. Wang, R. Xing, Y. Geng, Z. Zhai, P. Nakov, T. Baldwin, Loki: An open-source tool for fact verification, in: Proceedings of the 31st International Conference on Computational Lin- guistics: System Demonstrations, Association for Computational Lin- guistics, Abu Dhabi, UAE, 2025, pp. 28–36

    work page 2025

  8. [8]

    A. Asai, Z. Wu, Y. Wang, A. Sil, H. Hajishirzi, Self-rag: Learn- ing to retrieve, generate, and critique through self-reflection (2023). arXiv:2310.11511. URL https://arxiv.org/abs/2310.11511

    arXiv 2023

Show all 50 references
  1. [9]

    C. Li, J. Flanigan, RAC: efficient LLM factuality correction with re- trieval augmentation, CoRR abs/2410.15667 (2024). arXiv:2410.15667

    2024 arXiv

  2. [10]

    B. He, N. Chen, X. He, L. Yan, Z. Wei, J. Luo, Z. Ling, Retrieving, rethinking and revising: The chain-of-verification can improve retrieval augmented generation, in: Findings of the Association for Computa- tional Linguistics: EMNLP 2024, Miami, Florida, USA, November 12- 16,...

    2024

  3. [11]

    Y. Zhu, J. Si, Y. Zhao, H. Zhu, D. Zhou, Y. He, EXPLAIN, EDIT, GENERATE: Rationale-sensitive counterfactual data augmentation for multi-hop fact verification, in: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, Association for Computatio...

    2023

  4. [12]

    Subramanian, K

    S. Subramanian, K. Lee, Hierarchical Evidence Set Modeling for au- tomated fact extraction and verification, in: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), Association for Computational Linguistics, Online, 2020, pp. 7798–7809

    2020

  5. [13]

    Xiong, X

    W. Xiong, X. L. Li, S. Iyer, J. Du, P. Lewis, W. Y. Wang, Y. Mehdad, W.-t. Yih, S. Riedel, D. Kiela, et al., Answering complex open-domain questions with multi-hop dense retrieval, arXiv preprint arXiv:2009.12756 (2020)

    2020 arXiv

  6. [14]

    F. Xu, W. Shi, E. Choi, Recomp: Improving retrieval-augmented lms with compression and selective augmentation (2023). arXiv:2310.04408. 27

    2023 arXiv

  7. [15]

    Zhang, L

    C. Zhang, L. Zhang, D. Zhou, Causal walk: Debiasing multi-hop fact verification with front-door adjustment, CoRR abs/2403.02698 (2024). arXiv:2403.02698

    2024 arXiv

  8. [16]

    J. Si, Y. Zhu, D. Zhou, Exploring faithful rationale for multi-hop fact verification via salience-aware graph learning, CoRR abs/2212.01060 (2022). arXiv:2212.01060

    2022 arXiv

  9. [17]

    H. Wang, K. Shu, Explainable claim verification via knowledge-grounded reasoning with large language models, in: Findings of the Association for Computational Linguistics: EMNLP 2023, Association for Compu- tational Linguistics, Singapore, 2023, pp. 6288–6304

    2023

  10. [18]

    L. Pan, X. Lu, M.-Y. Kan, P. Nakov, QACheck: A demonstration sys- tem for question-guided multi-hop fact-checking, in: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Process- ing: System Demonstrations, Association for Computational Linguis- tics, ...

    2023

  11. [19]

    Zhang, W

    X. Zhang, W. Gao, Reinforcement retrieval leveraging fine-grained feed- back for fact checking news claims with black-box LLM, in: Proceedings of the 2024 Joint International Conference on Computational Linguis- tics, Language Resources and Evaluation, LREC/COLING 2024, 20-25 ...

    2024

  12. [20]

    S. Yan, J. Gu, Y. Zhu, Z. Ling, Corrective retrieval augmented genera- tion, CoRR abs/2401.15884 (2024). arXiv:2401.15884

    2024 arXiv

  13. [21]

    Z. Yue, H. Zeng, L. Shang, Y. Liu, Y. Zhang, D. Wang, Retrieval aug- mented fact verification by synthesizing contrastive arguments, in: Pro- ceedings of the 62nd Annual Meeting of the Association for Computa- tional Linguistics (Volume 1: Long Papers), ACL 2024, Bangkok, Thai...

    2024

  14. [22]

    B. Wang, J. Ma, H. Lin, Z. Yang, R. Yang, Y. Tian, Y. Chang, Ex- plainable fake news detection with large language model via defense among competing wisdom, in: Proceedings of the ACM on Web Confer- ence 2024, WWW 2024, Singapore, May 13-17, 2024, ACM, 2024, pp. 2452–2463. 28

    2024

  15. [23]

    Z. Shi, S. Zhang, W. Sun, S. Gao, P. Ren, Z. Chen, Z. Ren, Generate- then-ground in retrieval-augmented generation for multi-hop question answering, in: Proceedings of the 62nd Annual Meeting of the Associa- tion for Computational Linguistics (Volume 1: Long Papers), ACL 2024,...

    2024

  16. [24]

    Z. Xie, R. Xing, Y. Wang, J. Geng, H. Iqbal, D. Sahnan, I. Gurevych, P. Nakov, FIRE: fact-checking with iterative retrieval and verification, CoRR abs/2411.00784 (2024). arXiv:2411.00784

    2024

  17. [25]

    Gunjal, G

    A. Gunjal, G. Durrett, Molecular facts: Desiderata for decontextualiza- tion in LLM fact verification, in: Findings of the Association for Com- putational Linguistics: EMNLP 2024, Miami, Florida, USA, November 12-16, 2024, Association for Computational Linguistics, 2024, pp. 3...

    2024

  18. [26]

    Jiang, S

    Y. Jiang, S. Bordia, Z. Zhong, C. Dognin, M. Singh, M. Bansal, HoVer: A dataset for many-hop fact extraction and claim verification, in: Find- ings of the Association for Computational Linguistics: EMNLP 2020, Association for Computational Linguistics, Online, 2020, pp. 3441–3460

    2020

  19. [27]

    Z. Yang, J. Ma, H. Chen, H. Lin, Z. Luo, Y. Chang, A coarse-to-fine cascaded evidence-distillation neural network for explainable fake news detection, in: Proceedings of the 29th International Conference on Com- putational Linguistics, International Committee on Computational ...

    2022

  20. [28]

    liar, liar pants on fire

    W. Y. Wang, “liar, liar pants on fire”: A new benchmark dataset for fake news detection, in: Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), Association for Computational Linguistics, Vancouver, Canada, 2017, pp...

    2017

  21. [29]

    Alhindi, S

    T. Alhindi, S. Petridis, S. Muresan, Where is your evidence: Improv- ing fact-checking by justification modeling, in: Proceedings of the First Workshop on Fact Extraction and VERification (FEVER), Association for Computational Linguistics, Brussels, Belgium, 2018, pp. 85–90. 29

    2018

  22. [30]

    DeYoung, S

    J. DeYoung, S. Jain, N. F. Rajani, E. Lehman, C. Xiong, R. Socher, B. C. Wallace, ERASER: A benchmark to evaluate rationalized NLP models, in: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, Association for Computational Linguis- tics, ...

    2020

  23. [31]

    Paranjape, M

    B. Paranjape, M. Joshi, J. Thickstun, H. Hajishirzi, L. Zettlemoyer, An information bottleneck approach for controlling conciseness in ratio- nale extraction, in: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), Association for Co...

    2020

  24. [32]

    Glockner, I

    M. Glockner, I. Habernal, I. Gurevych, Why do you think that? explor- ing faithful sentence-level rationales without supervision, in: Findings of the Association for Computational Linguistics: EMNLP 2020, Asso- ciation for Computational Linguistics, Online, 2020, pp. 1080–1095

    2020

  25. [33]

    Carton, S

    S. Carton, S. Kanoria, C. Tan, What to learn, and how: To- ward effective learning from rationales, CoRR abs/2112.00071 (2021). arXiv:2112.00071

    2021 arXiv

  26. [34]

    J. Si, Y. Zhu, D. Zhou, Consistent multi-granular rationale extraction for explainable multi-hop fact verification, CoRR abs/2305.09400 (2023). arXiv:2305.09400

    2023 arXiv

  27. [35]

    Popat, S

    K. Popat, S. Mukherjee, A. Yates, G. Weikum, DeClarE: Debunking fake news and false claims using evidence-aware deep learning, in: Pro- ceedings of the 2018 Conference on Empirical Methods in Natural Lan- guage Processing, Association for Computational Linguistics, Brussels, B...

    2018

  28. [36]

    S. Jain, S. Wiegreffe, Y. Pinter, B. C. Wallace, Learning to faithfully rationalize by construction, in: Proceedings of the 58th Annual Meet- ing of the Association for Computational Linguistics, Association for Computational Linguistics, Online, 2020, pp. 4459–4473

    2020

  29. [37]

    S. M. Lundberg, S. Lee, A unified approach to interpreting model pre- dictions, in: Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, De- cember 4-9, 2017, Long Beach, CA, USA, 2017, pp. 4765–4774. 30

    2017

  30. [38]

    P. K. Mudrakarta, A. Taly, M. Sundararajan, K. Dhamdhere, Did the model understand the question?, in: Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), Association for Computational Linguistics, Melbourne, Australi...

    2018

  31. [39]

    De Cao, M

    N. De Cao, M. S. Schlichtkrull, W. Aziz, I. Titov, How do decisions emerge across layers in neural models? interpretation with differentiable masking, in: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), Association for Computa- t...

    2020

  32. [40]

    H. Chen, Y. Ji, Learning variational word masks to improve the inter- pretability of neural text classifiers, in: Proceedings of the 2020 Confer- ence on Empirical Methods in Natural Language Processing (EMNLP), Association for Computational Linguistics, Online, 2020, pp. 4236–4251

    2020

  33. [41]

    T. B. Brown, B. Mann, N. Ryder, M. Subbiah, J. Kaplan, P. Dhari- wal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell, S. Agarwal, A. Herbert-Voss, G. Krueger, T. Henighan, R. Child, A. Ramesh, D. M. Ziegler, J. Wu, C. Winter, C. Hesse, M. Chen, E. Sigler, M. Litwin, S. Gray, B...

    2020 arXiv

  34. [42]

    J. Wei, X. Wang, D. Schuurmans, M. Bosma, B. Ichter, F. Xia, E. Chi, Q. Le, D. Zhou, Chain-of-thought prompting elicits reasoning in large language models (2023). arXiv:2201.11903

    2023 arXiv

  35. [43]

    Zhang, W

    X. Zhang, W. Gao, Towards llm-based fact verification on news claims with a hierarchical step-by-step prompting method, in: Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Associati...

    2023

  36. [44]

    S. Yao, J. Zhao, D. Yu, N. Du, I. Shafran, K. Narasimhan, Y. Cao, React: Synergizing reasoning and acting in language models (2023). arXiv:2210.03629. 31

    2023 arXiv

  37. [45]

    W. Y. Wang, ”liar, liar pants on fire”: A new benchmark dataset for fake news detection, in: Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, ACL 2017, Vancouver, Canada, July 30 - August 4, Volume 2: Short Papers, Association for Comput...

    2017

  38. [46]

    Rashkin, E

    H. Rashkin, E. Choi, J. Y. Jang, S. Volkova, Y. Choi, Truth of varying shades: Analyzing language in fake news and political fact-checking, in: Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017, Copenhagen, Denmark, September 9-...

    2017

  39. [47]

    J. Ma, W. Gao, S. R. Joty, K. Wong, Sentence-level evidence embed- ding for claim verification with hierarchical attention networks, in: Pro- ceedings of the 57th Conference of the Association for Computational Linguistics, ACL 2019, Florence, Italy, July 28- August 2, 2019, V...

    2019

  40. [48]

    Kotonya, F

    N. Kotonya, F. Toni, Explainable automated fact-checking for pub- lic health claims, in: Proceedings of the 2020 Conference on Empir- ical Methods in Natural Language Processing, EMNLP 2020, Online, November 16-20, 2020, Association for Computational Linguistics, 2020, pp. 7740–7754

    2020

  41. [49]

    Atanasova, J

    P. Atanasova, J. G. Simonsen, C. Lioma, I. Augenstein, Generating fact checking explanations, in: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, ACL 2020, Online, July 5-10, 2020, Association for Computational Linguistics, 2020, pp. 7352– 7364

    2020

  42. [50]

    Z. Yang, J. Ma, H. Chen, H. Lin, Z. Luo, Y. Chang, A coarse-to-fine cascaded evidence-distillation neural network for explainable fake news detection, in: Proceedings of the 29th International Conference on Com- putational Linguistics, COLING 2022, Gyeongju, Republic of Korea,...

    2022

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.