Pith. sign in

REVIEW 4 major objections 8 minor 76 references

Wisdom in Unity: The Role of Multilingual Training in Figurative Language Identification in Proverbs

T0 review · 4 major / 8 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Adding translated proverb data improves figurative-language identification, with performance plateauing near 50% of the multilingual training set.

desk verdict Genuinely new annotation framework for proverbs, but the central findings are measured against labels the authors' own weak-supervision models produced, so the curves describe agreement with the labeler. read the letter →

arxiv 2608.08090 v1 pith:4AIVX4HX submitted 2026-08-08 cs.CL

Rama Alomair , Remas Alsubaie , Walaa Saifalislam , Rima Alsonbul , Mona Alnajjar , Razan Aldossari , Haya Alibrahim , Abeer Aldayel This is my paper · ORCID
classification cs.CL
keywords figurativelanguageidentificationmultilingualsupervisionproverbsmultidimensionalannotationframeworkculture-specificmeaningweakcross-lingualtransferMacro-F1evaluation
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

The paper asks whether translated multilingual training data helps a model recognize whether a proverb is figurative, and how much translated data is actually needed. Using 742 proverb concepts across 6,787 instances in Arabic, English, French, German, Russian, Japanese, and Spanish, it evaluates five models (logistic regression, multilingual BERT, XLM-R Large, and two instruction-tuned LLMs) at 0%, 10%, 50%, and 100% multilingual supervision, measuring Macro-F1. The central claim is that about 50% of the translated training data is enough to reach near-optimal figurative-identification performance, and that the rarest figurative form, Culture-Specific, gains the most from multilingual supervision. The paper also introduces a four-part annotation framework for proverbs: Metaphorical, Moral/Advisory, Cause-Effect, and Culture-Specific. If correct, this would mean multilingual figurative understanding should be trained and evaluated per figurative form rather than treated as one metaphor-centric task.

What carries the argument

The machinery is a concept-level, multidimensional annotation framework grounded in Relevance Theory, which characterizes each proverb by four non-exclusive figurative forms: Metaphorical, Moral/Advisory, Cause-Effect, and Culture-Specific. A weak-supervision pipeline scales a human-validated seed of 83 proverb concepts to the full 6,787-instance dataset: XLM-R Base produces the binary figurative/literal labels, E5 embeddings with logistic regression produce the four fine-grained labels (reaching a Macro-F1 of 0.36 on that task), and labels are propagated from each source proverb to all of its translations. The framework carries the argument by making it possible to compute supervision curves per figurative form rather than only for a single figurative-versus-literal label, and the Jaccard-overlap analysis shows Culture-Specific is the most distinct form, which is the precondition for interpreting its large gains as a distinct effect.

What would settle it

Have human annotators label a held-out sample of the test proverbs in all seven languages with the same four-form guideline, then re-run the 0%, 10%, 50%, and 100% supervision curves on those human labels; the central claim would be wrong if the plateau is not near 50% or if Culture-Specific is not the largest gainer.

Watch

Extended reading notes

Core claim

The paper's central discovery is that multilingual supervision does not help all figurative meanings equally, and its signature result is the rarest form benefiting most. In a controlled comparison, adding translated multilingual training instances to a fixed source-language training set improves binary figurative-identification Macro-F1 for every model, and the curve flattens around 50% of the translated data: moving from 50% to 100% produces smaller gains than moving from 0% to 50%. Subgroup analyses show Culture-Specific proverbs, the least frequent form and the one with the lowest overlap with the other three, post the largest F1 gains under multilingual supervision, for example from 0.481 to 0.808 for logistic regression and from 0.278 to 0.796 for XLM-R Large, while the Literal/None subgroup sometimes declines. The paper further claims that training on all four figurative forms together gives the strongest overall performance, and that Moral/Advisory and Culture-Specific forms contribute most to instruction-tuned LLMs. It emphasizes that these results hold at the distributional level: the prevalence of the four forms is preserved across languages, but the paper does not claim every individual translation preserves the same cultural interpretation.

Load-bearing premise

The load-bearing premise is that the automatically generated labels used for training and testing are accurate enough: most of the dataset is labeled by a weak-supervision model, the fine-grained labeler reaches only 0.36 Macro-F1, and every translation of a source proverb receives the same label, so a systematic labeling bias would make the 50% plateau and the culture-specific gains artifacts of the labelers rather than properties of proverbs.

Editorial extensions

If this is right

  • Adding translated multilingual training data significantly improves binary figurative identification in proverbs, with all five models fixing more test errors than they introduce when supervision moves from 0% to 100%.
  • About 50% of the translated training data is enough to approach the performance ceiling; doubling to 100% adds only small gains, so the multilingual benefit is not simply proportional to data volume.
  • The four figurative forms are not equally transferable: Metaphorical gives the strongest gains for mBERT and Aya, Moral/Advisory for Qwen2.5-3B, and Cause-Effect for XLM-R Large, but combining all forms yields the best overall performance.
  • Culture-Specific, the rarest and most distinct form, gains the most from multilingual supervision, suggesting that multilingual data differentially helps figurative knowledge that is hardest to learn from shared cues.

Reading between the lines

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

  • Editorial inference: If the Culture-Specific gain is driven by its low overlap with the other forms, then multilingual supervision may help most for any rare, semantically distinct figurative category; a testable prediction is that irony, allusion, or historically grounded sayings would show similar gain curves.
  • Editorial inference: The paper's distributional-preservation result does not establish that individual translations keep the same cultural meaning, so a direct concept-level comparison of aligned translations with human judgments would be the natural next experiment to confirm the culture-specific effect is real grounding rather than label propagation.
  • Editorial inference: The 50% plateau gives a practical data-collection target for proverb-style figurative tasks: collecting roughly half of a translated training set may capture most of the benefit, letting future benchmarks allocate annotation effort to more languages or forms instead.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 8 minor

Summary. The paper studies the effect of progressively adding translated multilingual training data on binary figurative-language identification in proverbs. It introduces a four-way fine-grained annotation framework (Metaphorical, Moral/Advisory, Cause–Effect, Culture-Specific) and evaluates five model families across seven languages at 0%, 10%, 50%, and 100% multilingual supervision. The authors report that performance saturates around 50% of the translated data, that combining all figurative forms gives the strongest supervision signal, that the least frequent form (Culture-Specific) shows the largest gains, and that the proposed forms are distributionally stable across languages. A controlled instance-selection ablation isolates the contribution of each fine-grained form. The dataset and annotations are to be released upon publication.

Significance. If the reported trends held on human-validated labels, the 50% saturation result and the Culture-Specific gains would be practically valuable for multilingual data selection and annotation. The proposed multidimensional framework is a reasonable diagnostic lens that goes beyond metaphor-centric taxonomies, and the paper includes a commendable breadth of models, controlled supervision levels, and statistical significance testing. However, the significance is heavily conditional because the evaluation relies on weakly supervised silver labels generated by the same pipeline that labels the training data, and the fine-grained labeler has only 0.36 Macro-F1 on the human-validated seed. The claimed central findings are therefore not yet established.

major comments (4)
  1. [Section 3, Weak Supervision Labeling across Languages] The test set used for all reported experiments is labeled by the weak-supervision pipeline rather than by human annotation. Binary labels come from XLM-R Base, and fine-grained form labels come from E5+LR, which reaches only 0.3640 Macro-F1 on the human-validated seed (Table 11). The paper's own Ethical Considerations state that these labels 'should not be regarded as human gold annotations.' Consequently, Figures 2 and 4 and Tables 7-8 measure agreement with the weak labeler, not figurative-language identification accuracy. The central claims about 50% saturation and Culture-Specific gains are unsupported until the same analyses are shown on a human-labeled test set.
  2. [Appendix A, label propagation] The pipeline propagates each source proverb's label to all translations sharing the same source ID. This assumes instance-level figurative equivalence across languages, yet the paper's own Table 16 and Discussion acknowledge that translations of the same concept can use different cultural imagery and that categorical consistency does not establish instance-level preservation. Because the test set contains no per-language human ground truth, the per-language results in Figure 3 and Table 12 are confounded by propagated label noise, and the cross-lingual transfer conclusion is not separately validated.
  3. [Section 4.2 and Figure 4] The fine-grained instance-selection ablation uses the same E5+LR labels both to select training subsets and to define the test subgroups. This creates a shared-bias confound: when the labeler is biased for a form, training on silver-labeled instances of that form can improve agreement with the same labeler on the test subgroup without improving genuine figurative identification. The Culture-Specific result is especially vulnerable because the fine-grained labeler has low Macro-F1 overall (0.3640) and this rare form is acknowledged as the least prevalent (Table 2). The authors should report the human-validated 83-concept subset separately to demonstrate that the same trends hold.
  4. [Section 5, RQ1 saturation claim] The statement that performance 'generally stabilizes after incorporating approximately 50% of the translated multilingual training data' is not formally tested. The McNemar analyses in Table 13 compare only 0% versus 100% supervision; no test compares 50% versus 100% or quantifies the gain from 50% to 100% across models. Without such a comparison, the saturation claim is an informal reading of Figure 2 rather than a statistically supported finding.
minor comments (8)
  1. [Section 1, Contributions] The word 'languges' is a typo and should be 'languages.'
  2. [Section 4.2] The phrase 'biary figritve detection' contains two typos; it should be 'binary figurative detection.'
  3. [Section 5] 'Macnammer significant test' should be 'McNemar significance test.'
  4. [Table 5] The annotation guideline is internally inconsistent: the Output row says 'A single-select label is assigned' while the Decision rule says annotators 'choose tall representative labels' and labels may overlap. The multi-label nature should be stated consistently.
  5. [Ethical Considerations] 'recrutment' is a typo for 'recruitment.'
  6. [References] Several references use the placeholder style 'and 1 others' (e.g., Benamara et al., Chuang et al., Ghosh et al., Lai et al., Liu et al.); these should be replaced with the full author lists.
  7. [Figure 4] The four subplots lack axis labels, and the caption references (a)-(d) but the displayed figure does not label the subplots clearly.
  8. [Section 5] Table 12 is not cited in the main text; it should be referenced where per-language results are discussed.

Circularity Check

1 steps flagged · score 6.0 of 10

Weakly supervised, source-propagated silver labels define both training and test, so the reported 50% saturation and Culture-Specific gains measure agreement with the paper's own weak labelers rather than independent figurative-language identification.

  1. fitted input called prediction [Appendix A, 'Weak Supervision Figurative binary Labeling'; Section 3, 'Weak Supervision Labeling across languages'; Ethical Considerations]
    "The generated labels were then propagated to all translations sharing the same source proverb identifier, ensuring that conceptually equivalent proverbs received consistent labels regardless of language."

    The paper's train and test sets are both drawn from a corpus whose labels were generated by weak-supervision models: XLM-R Base for binary figurative-versus-literal and E5+LR for the four fine-grained forms. These silver labels were propagated from each source proverb to every translation sharing the same source identifier, so the test set contains no per-language human ground truth. The central results (Figures 2 and 4, Tables 7-8) therefore report how well the five trained models reproduce the precomputed weak-label decisions.

full rationale

The paper's main contribution is an empirical study of how progressively added multilingual supervision affects figurative-proverb identification. On inspection, the evaluation target is not an independent human-labeled test set: binary labels are produced by XLM-R Base, fine-grained form labels by E5+LR, and both are propagated from source proverbs to all translations via the same source identifier. The paper's own Ethical Considerations disclaims gold-annotation status, and the Discussion concedes that categorical consistency does not establish instance-level cultural preservation. Consequently, the reported saturation curve and Culture-Specific gain measure how well the trained classifiers reproduce the authors' weak labelers, which is a partial circularity: the 'prediction' is agreement with the label-generating pipeline rather than with external figurative-language judgments. I do not count the self-citation of the proverb dataset (Alshaalan et al. 2026, with overlapping authorship) as load-bearing because the dataset is only raw material; the weak-labeling pipeline and four-form annotation scheme are new. The +All-target condition in Table 4 is also confounded with training-set size, but that is an experimental-design concern rather than a circularity. Overall score 6 reflects that the central empirical findings reduce, by construction, to behavior on self-generated silver labels.

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

The paper's analysis rests on a small set of domain assumptions about proverb annotation and on the reliability of weakly supervised labels; there are no fitted free parameters in the usual sense, but the supervision levels and the 600-instance ablation size are design choices.

assumptions (4)
  • domain assumption The four figurative forms (Metaphorical, Moral/Advisory, Cause-Effect, Culture-Specific) are complementary, meaningful, and sufficient dimensions for characterizing proverb figurativeness.
    Introduced in Section 3 and used to define all subgroup analyses; inter-annotator agreement is moderate (kappa 0.42-0.62), so the categories are not objectively forced.
  • domain assumption Weakly supervised silver labels are accurate enough to serve as evaluation ground truth.
    Binary labels come from XLM-R Base applied to the proverb set; fine-grained labels come from E5+LR with only 0.36 Macro-F1 on the human seed. The paper acknowledges this in the Limitations section.
  • domain assumption Labels propagate unchanged from a source proverb to all its translations.
    Section 3: 'generated labels were then propagated to all translations sharing the same source proverb identifier'. This assumes concept-level figurative meaning is language-invariant, which the Discussion says is not established at instance level.
  • domain assumption English is the fixed source language in the progressive supervision design.
    The 0% condition uses only source-language training instances; per-language 0% Macro-F1 values near 0.33 for non-English languages confirm that zero-shot cross-lingual performance is near chance.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Wisdom in Unity: The Role of Multilingual Training in Figurative Language Identification in Proverbs." pith.science (2026). https://pith.science/paper/4AIVX4HX

@misc{pith2026260808090,
  author       = {Pith},
  title        = {Pith review of: Wisdom in Unity: The Role of Multilingual Training in Figurative Language Identification in Proverbs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4AIVX4HX}},
  note         = {Machine review of arXiv:2608.08090}
}
read the original abstract

Although multilingual approaches to figurative language identification are not new, the shift beyond language homogeneous training data requires a clearer understanding of the contribution of translated multilingual supervision. We examine this question using 742 proverb concepts across 6,787 translated instances in seven languages. We evaluate five models, including multilingual encoders and instruction tuned LLMs, under progressively increasing levels of multilingual supervision. Moreover, we introduce a multidimensional annotation framework for proverbs that characterizes them through four complementary figurative forms: Metaphorical, Moral/Advisory, Cause-Effect, and Culture Specific. Our findings show that approximately 50% of the translated multilingual training data is sufficient to achieve near-optimal figurative language identification performance. We further show that combining diverse figurative forms yields the strongest overall performance. A notable finding is that the least frequent figurative form, Culture Specific, exhibits the largest performance gains under multilingual supervision. Furthermore, the Moral/Advisory and Culture Specific forms contribute most to the performance of instruction-tuned LLMs on figurative language identification. These findings motivate multilingual figurative language identification to move beyond metaphor-centric taxonomies toward concept level multidimensional frameworks that explicitly model complementary forms of figurative meaning.

Figures

Figures reproduced from arXiv: 2608.08090 by the authors.

Figure 1
Figure 1. Pairwise Jaccard similarity between the four [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Overall Macro-F1 across increasing levels of [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 4
Figure 4. Test-set subgroup F1 across multilingual su [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Prevalence of the four fine-grained figurative [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

76 extracted references · 67 canonical work pages

  1. [1]

    Aho and Jeffrey D

    Alfred V. Aho and Jeffrey D. Ullman , title =. 1972

    1972

  2. [2]

    Publications Manual , year = "1983", publisher =

    1983

  3. [3]

    Chandra and Dexter C

    Ashok K. Chandra and Dexter C. Kozen and Larry J. Stockmeyer , year = "1981", title =. doi:10.1145/322234.322243

    arXiv 1981

  4. [4]

    Scalable training of

    Andrew, Galen and Gao, Jianfeng , booktitle=. Scalable training of

  5. [5]

    Dan Gusfield , title =. 1997

    1997

  6. [6]

    Tetreault , title =

    Mohammad Sadegh Rasooli and Joel R. Tetreault , title =. Computing Research Repository , volume =. 2015 , url =

    2015

  7. [7]

    A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data , Volume =

    Ando, Rie Kubota and Zhang, Tong , Issn =. A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data , Volume =. Journal of Machine Learning Research , Month = dec, Numpages =

  8. [9]

    2023 , url =

    Lai, Huiyuan and others , booktitle =. 2023 , url =

    work page 2023

Show all 76 references
  1. [10]

    arXiv preprint , year=

    MAPS: Multilingual Annotated Proverbs Dataset , author=. arXiv preprint , year=

  2. [11]

    Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

    So Many Phrases, So Little Time: On Memorization in Multilingual Models. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023

    2023

  3. [12]

    Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics

    How Multilingual is Multilingual BERT?. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. 2019

    2019

  4. [13]

    2023 , url =

    MEMPHIS-data , author =. 2023 , url =

    2023

  5. [14]

    Proceedings of the 16th International Workshop on Semantic Evaluation (

    Madabushi, Harish Tayyar and Gow-Smith, Edward and Garcia, Marcos and Scarton, Carolina and Idiart, Marco and Villavicencio, Aline , title =. Proceedings of the 16th International Workshop on Semantic Evaluation (. 2022 , pages =

    2022

  6. [15]

    Proceedings of the 16th International Workshop on Semantic Evaluation (

    Yamaguchi, Akira and Hirao, Wakako and Nagata, Masaaki , title =. Proceedings of the 16th International Workshop on Semantic Evaluation (. 2022 , pages =

    2022

  7. [16]

    Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing , year =

    Chakrabarty, Tuhin and Saakyan, Arkadiy and Ghosh, Debanjan and Muresan, Smaranda , title =. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing , year =

    2022

  8. [17]

    International Journal of Market Research , volume =

    Sarsam, Samer Muthana and Al-Samarraie, Hosam and Alzahrani, Ahmed Ibrahim and Wright, Bianca , title =. International Journal of Market Research , volume =. 2020 , doi =

    2020

  9. [18]

    2022 , url =

    Novic, Lisa Irene , title =. 2022 , url =

    2022

  10. [19]

    Multilingual Proverb Translation Dataset , year =

  11. [20]

    Relevance: Communication and cognition

    Sperber, D and Wilson, D. Relevance: Communication and cognition

  12. [21]

    https://osf.io/rckmq/overview?view_only=13656c37ad1243058535406191ab0548 Multilingual proverb translation dataset

    2026. https://osf.io/rckmq/overview?view_only=13656c37ad1243058535406191ab0548 Multilingual proverb translation dataset

    2026

  13. [22]

    Saeed Almheiri, Bilal Elbouardi, Salsabila Zahirah Pranida, Irina Nikishina, Ashwath Rao, Parameswari Krishnamurthy, Muhammad Cendekia Airlangga, Rifo Ahmad Genadi, Nguyen Phan Gia Bao, Amir Hossein Yari, Hawau Olamide Toyin, Nurdaulet Mukhituly, Mena Attia, Besher Hassan, Ahm...

    2026

  14. [23]

    Jory Alshaalan, Haya Albaker, Abeer Aldayel, Aljawharah Alabdullatif, and Rehab Alahmadi. 2026. Same lesson, different story: Cross-lingual reconstruction of cultural narratives in large language models. arXiv [cs.CL]

    2026

  15. [24]

    Israa Alsiyat and Scott S L Piao. 2020. Metaphorical expressions in automatic arabic sentiment analysis. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 4911--4916

    2020

  16. [25]

    Naveen Badathala, Abisek Rajakumar Kalarani, Tejpalsingh Siledar, and Pushpak Bhattacharyya. 2023. A match made in heaven: A multi-task framework for hyperbole and metaphor detection. In Findings of the Association for Computational Linguistics: ACL 2023, pages 388--401, Strou...

    2023

  17. [26]

    Zouheir Benamara and 1 others. 2023. https://github.com/ZOUHEIRBN/MEMPHIS-data Memphis-data

    2023

  18. [27]

    Tuhin Chakrabarty, Yejin Choi, and Vered Shwartz. 2022 a . It’s not rocket science: Interpreting figurative language in narratives. Trans. Assoc. Comput. Linguist., 10:589--606

    2022

  19. [28]

    Tuhin Chakrabarty, Arkadiy Saakyan, Debanjan Ghosh, and Smaranda Muresan. 2022 b . https://aclanthology.org/2022.emnlp-main.481/ FLUTE : Figurative language understanding through textual explanations . In Proceedings of the 2022 Conference on Empirical Methods in Natural Langu...

    2022

  20. [29]

    Yung-Sung Chuang and 1 others. 2023. https://aclanthology.org/2023.emnlp-main.406/ So many phrases, so little time: On memorization in multilingual models . In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

    2023

  21. [30]

    Alexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzmán, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Unsupervised cross-lingual representation learning at scale. CoRR, abs/1911.02116

    2019 arXiv

  22. [31]

    John Dang, Shivalika Singh, Daniel D'souza, Arash Ahmadian, Alejandro Salamanca, Madeline Smith, Aidan Peppin, Sungjin Hong, Manoj Govindassamy, Terrence Zhao, Sandra Kublik, Meor Amer, Viraat Aryabumi, Jon Ander Campos, Yi-Chern Tan, Tom Kocmi, Florian Strub, Nathan Grinsztaj...

    2024

  23. [32]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. BERT : Pre-training of deep bidirectional transformers for language understanding. CoRR, abs/1810.04805

    2018 arXiv

  24. [33]

    Saptarshi Ghosh and 1 others. 2023. https://arxiv.org/abs/2305.16171 Figbench: A benchmark for figurative language understanding . arXiv preprint arXiv:2305.16171

    2023 arXiv

  25. [34]

    Gamze Goren and Carlo Strapparava. 2024. Context matters: Enhancing metaphor recognition in proverbs. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 3825--3830

    2024

  26. [35]

    Faezeh Hosseini, Mohammadali Yousefzadeh, and Yadollah Yaghoobzadeh. 2026. FFE -hallu: Hallucinations in fixed figurative expressions: A benchmark of idioms and proverbs in the persian language. In Proceedings of the 19th Conference of the European Chapter of the Association f...

    2026

  27. [36]

    H Hrisztova-Gotthardt and M A Varga. 2015. Introduction to paremiology: A comprehensive guide to proverb studies. Sciendo, Berlin, Germany

    2015

  28. [37]

    Anna Hülsing and Sabine Schulte Im Walde. 2024. Cross-lingual metaphor detection for low-resource languages. In Proceedings of the 4th Workshop on Figurative Language Processing (FigLang 2024), pages 22--34, Stroudsburg, PA, USA. Association for Computational Linguistics

    2024

  29. [38]

    Hyeju Jang, Keith Maki, Eduard Hovy, and Carolyn Rose. 2017. Finding structure in figurative language: Metaphor detection with topic-based frames. In Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue, pages 320--330, Stroudsburg, PA, USA. Association for...

    2017

  30. [39]

    Anubha Kabra, Emmy Liu, Simran Khanuja, Alham Fikri Aji, Genta Winata, Samuel Cahyawijaya, Anuoluwapo Aremu, Perez Ogayo, and Graham Neubig. 2023. Multi-lingual and multi-cultural figurative language understanding. In Findings of the Association for Computational Linguistics: ...

    2023

  31. [40]

    Paria Khoshtab, Danial Namazifard, Mostafa Masoudi, Ali Akhgary, Samin Mahdizadeh Sani, and Yadollah Yaghoobzadeh. 2025. Comparative study of multilingual idioms and similes in large language models. In Proceedings of the 31st International Conference on Computational Linguist...

    2025

  32. [41]

    Huiyuan Lai, Antonio Toral, and Malvina Nissim. 2023 a . Multilingual multi-figurative language detection. In Findings of the Association for Computational Linguistics: ACL 2023, pages 9254--9267, Stroudsburg, PA, USA. Association for Computational Linguistics

    2023

  33. [42]

    Huiyuan Lai and 1 others. 2023 b . https://github.com/laihuiyuan/MMFLD/tree/main/data MMFLD : A multilingual multi-figure figurative language dataset . In Proceedings of EMNLP

    2023

  34. [43]

    Chen Liu, Fajri Koto, Timothy Baldwin, and Iryna Gurevych. 2024. Are multilingual LLMs culturally-diverse reasoners? an investigation into multicultural proverbs and sayings. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computation...

    2024

  35. [44]

    Liu and 1 others

    Nelson F. Liu and 1 others. 2023. https://github.com/UKPLab/maps Maps: Multilingual annotated proverbs dataset . arXiv preprint

    2023

  36. [45]

    Harish Tayyar Madabushi, Edward Gow-Smith, Marcos Garcia, Carolina Scarton, Marco Idiart, and Aline Villavicencio. 2022. https://aclanthology.org/2022.semeval-1.13/ SemEval -2022 task 2: Multilingual idiomaticity detection and sentence embedding . In Proceedings of the 16th In...

    2022

  37. [46]

    Samar Mohamed Magdy, Sang Yun Kwon, Fakhraddin Alwajih, Safaa Taher Abdelfadil, Shady Shehata, and Muhammad Abdul-Mageed. 2025. JAWAHER : A multidialectal dataset of arabic proverbs for LLM benchmarking. In Proceedings of the 2025 Conference of the Nations of the Americas Chap...

    2025

  38. [47]

    Vlad Niculae and Cristian Danescu-Niculescu-Mizil. 2014. Brighter than gold: Figurative language in user generated comparisons. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 2008--2018, Stroudsburg, PA, USA. Associatio...

    2014

  39. [48]

    Soyoung Oh, Xinting Huang, Mathis Pink, Michael Hahn, and Vera Demberg. 2026. Tug-of-war between idioms’ figurative and literal interpretations in LLMs . In Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: L...

    2026

  40. [49]

    Qwen Team . 2024. Qwen2 .5: A party of foundation models

    2024

  41. [50]

    D Sperber and D Wilson. 1986. Relevance: Communication and cognition. 142

    1986

  42. [51]

    Minghan Wang, Viet Thanh Pham, Farhad Moghimifar, and Thuy-Trang Vu. 2025. Proverbs run in pairs: Evaluating proverb translation capability of large language model. In Findings of the Association for Computational Linguistics: ACL 2025, pages 1646--1662, Stroudsburg, PA, USA. ...

    2025

  43. [52]

    Akira Yamaguchi, Wakako Hirao, and Masaaki Nagata. 2022. https://aclanthology.org/2022.semeval-1.15/ Hitachi at SemEval -2022 task 2: On the effectiveness of pre-trained language model for multilingual idiomaticity detection . In Proceedings of the 16th International Workshop ...

    2022

  44. [53]

    Gözde Özbal, Carlo Strapparava, Serra Sinem Tekiroglu, and Daniele Pighin. 2016. Learning to identify metaphors from a corpus of proverbs. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 2060--2065, Stroudsburg, PA, USA. Associa...

    2016

  45. [54]

    Learning to identify metaphors from a corpus of proverbs

    Özbal, Gözde and Strapparava, Carlo and Tekiroglu, Serra Sinem and Pighin, Daniele. Learning to identify metaphors from a corpus of proverbs. Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing

    2016

  46. [55]

    It’s not rocket science: Interpreting figurative language in narratives

    Chakrabarty, Tuhin and Choi, Yejin and Shwartz, Vered. It’s not rocket science: Interpreting figurative language in narratives. Trans. Assoc. Comput. Linguist

  47. [56]

    Multi-lingual and Multi-cultural Figurative Language Understanding

    Kabra, Anubha and Liu, Emmy and Khanuja, Simran and Aji, Alham Fikri and Winata, Genta and Cahyawijaya, Samuel and Aremu, Anuoluwapo and Ogayo, Perez and Neubig, Graham. Multi-lingual and Multi-cultural Figurative Language Understanding. Findings of the Association for Computa...

    2023

  48. [57]

    Qwen2 .5: A Party of Foundation Models

    Qwen Team. Qwen2 .5: A Party of Foundation Models

  49. [58]

    Proverbs run in pairs: Evaluating proverb translation capability of large language model

    Wang, Minghan and Pham, Viet Thanh and Moghimifar, Farhad and Vu, Thuy-Trang. Proverbs run in pairs: Evaluating proverb translation capability of large language model. Findings of the Association for Computational Linguistics: ACL 2025

    2025

  50. [59]

    Are multilingual LLMs culturally-diverse reasoners? An investigation into multicultural proverbs and sayings

    Liu, Chen and Koto, Fajri and Baldwin, Timothy and Gurevych, Iryna. Are multilingual LLMs culturally-diverse reasoners? An investigation into multicultural proverbs and sayings. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computation...

    2024

  51. [60]

    Anthropomimetic uncertainty: What verbalized uncertainty in language models is missing

    Ulmer, Dennis and Lorson, Alexandra and Titov, Ivan and Hardmeier, Christian. Anthropomimetic uncertainty: What verbalized uncertainty in language models is missing. arXiv [cs.CL]

  52. [61]

    Same lesson, different story: Cross-lingual reconstruction of cultural narratives in large language models

    Alshaalan, Jory and Albaker, Haya and Aldayel, Abeer and Alabdullatif, Aljawharah and Alahmadi, Rehab. Same lesson, different story: Cross-lingual reconstruction of cultural narratives in large language models. arXiv [cs.CL]

  53. [62]

    Metaphorical Expressions in Automatic Arabic Sentiment Analysis

    Alsiyat, Israa and Piao, Scott S L. Metaphorical Expressions in Automatic Arabic Sentiment Analysis. Proceedings of the Twelfth Language Resources and Evaluation Conference

  54. [63]

    Aya Expanse: Combining Research Breakthroughs for a New Multilingual Frontier

    Dang, John and Singh, Shivalika and D'souza, Daniel and Ahmadian, Arash and Salamanca, Alejandro and Smith, Madeline and Peppin, Aidan and Hong, Sungjin and Govindassamy, Manoj and Zhao, Terrence and Kublik, Sandra and Amer, Meor and Aryabumi, Viraat and Campos, Jon Ander and ...

  55. [64]

    A match made in heaven: A multi-task framework for hyperbole and metaphor detection

    Badathala, Naveen and Rajakumar Kalarani, Abisek and Siledar, Tejpalsingh and Bhattacharyya, Pushpak. A match made in heaven: A multi-task framework for hyperbole and metaphor detection. Findings of the Association for Computational Linguistics: ACL 2023

    2023

  56. [65]

    Multilingual multi-figurative language detection

    Lai, Huiyuan and Toral, Antonio and Nissim, Malvina. Multilingual multi-figurative language detection. Findings of the Association for Computational Linguistics: ACL 2023

    2023

  57. [66]

    Cross-lingual metaphor detection for low-resource languages

    Hülsing, Anna and Schulte Im Walde, Sabine. Cross-lingual metaphor detection for low-resource languages. Proceedings of the 4th Workshop on Figurative Language Processing (FigLang 2024)

    2024

  58. [67]

    Introduction to paremiology: A comprehensive guide to proverb studies

    Hrisztova-Gotthardt, H and Varga, M A. Introduction to paremiology: A comprehensive guide to proverb studies

  59. [68]

    JAWAHER : A Multidialectal Dataset of Arabic Proverbs for LLM Benchmarking

    Magdy, Samar Mohamed and Kwon, Sang Yun and Alwajih, Fakhraddin and Abdelfadil, Safaa Taher and Shehata, Shady and Abdul-Mageed, Muhammad. JAWAHER : A Multidialectal Dataset of Arabic Proverbs for LLM Benchmarking. Proceedings of the 2025 Conference of the Nations of the Ameri...

    2025

  60. [69]

    FFE -hallu: Hallucinations in fixed figurative expressions: A benchmark of idioms and proverbs in the Persian language

    Hosseini, Faezeh and Yousefzadeh, Mohammadali and Yaghoobzadeh, Yadollah. FFE -hallu: Hallucinations in fixed figurative expressions: A benchmark of idioms and proverbs in the Persian language. Proceedings of the 19th Conference of the European Chapter of the Association for C...

  61. [70]

    Comparative Study of Multilingual Idioms and Similes in Large Language Models

    Khoshtab, Paria and Namazifard, Danial and Masoudi, Mostafa and Akhgary, Ali and Sani, Samin Mahdizadeh and Yaghoobzadeh, Yadollah. Comparative Study of Multilingual Idioms and Similes in Large Language Models. Proceedings of the 31st International Conference on Computational ...

  62. [71]

    Context Matters: Enhancing Metaphor Recognition in Proverbs

    Goren, Gamze and Strapparava, Carlo. Context Matters: Enhancing Metaphor Recognition in Proverbs. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)

    2024

  63. [72]

    Multilingual idioms in sentences and conversations across high-, medium-, and low-resource languages

    Almheiri, Saeed and Elbouardi, Bilal and Pranida, Salsabila Zahirah and Nikishina, Irina and Rao, Ashwath and Krishnamurthy, Parameswari and Airlangga, Muhammad Cendekia and Genadi, Rifo Ahmad and Bao, Nguyen Phan Gia and Yari, Amir Hossein and Toyin, Hawau Olamide and Mukhitu...

  64. [73]

    Finding structure in figurative language: Metaphor detection with topic-based frames

    Jang, Hyeju and Maki, Keith and Hovy, Eduard and Rose, Carolyn. Finding structure in figurative language: Metaphor detection with topic-based frames. Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue

  65. [74]

    Brighter than gold: Figurative language in user generated comparisons

    Niculae, Vlad and Danescu-Niculescu-Mizil, Cristian. Brighter than gold: Figurative language in user generated comparisons. Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP)

    2014

  66. [75]

    Tug-of-war between idioms’ figurative and literal interpretations in LLMs

    Oh, Soyoung and Huang, Xinting and Pink, Mathis and Hahn, Michael and Demberg, Vera. Tug-of-war between idioms’ figurative and literal interpretations in LLMs. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1...

  67. [76]

    BERT : Pre-training of Deep Bidirectional Transformers for Language Understanding

    Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina. BERT : Pre-training of Deep Bidirectional Transformers for Language Understanding. CoRR

  68. [77]

    Unsupervised Cross-lingual Representation Learning at Scale

    Conneau, Alexis and Khandelwal, Kartikay and Goyal, Naman and Chaudhary, Vishrav and Wenzek, Guillaume and Guzmán, Francisco and Grave, Edouard and Ott, Myle and Zettlemoyer, Luke and Stoyanov, Veselin. Unsupervised Cross-lingual Representation Learning at Scale. CoRR

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.