Pith. sign in

REVIEW 3 major objections 8 minor 42 references

Large Vision-Language Models for Knowledge-Grounded Data Annotation of Memes

T0 review · 3 major / 8 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read Fine-tuning CLIP on GPT-4o meme captions lifts retrieval recall by up to 17 points.

desk verdict Useful dataset, but the central retrieval claim is under-supported: the MemeCap gain may be template memorization, and the paper doesn't provide the split to rule it out. read the letter →

arxiv 2501.13851 v1 pith:PT7T23CL submitted 2025-01-23 cs.LG

classification cs.LG
keywords CM50meme-textretrievalCLIPfine-tuningGPT-4oannotationmemecomprehensiontemplate-baseddatasetliterarydevicesvision-languagemodels
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 argues that large vision-language models can replace human annotators in building training data for meme understanding at scale. It introduces CM50, 33,172 ImgFlip memes across 50 templates, annotated by GPT-4o with knowledge of each template's context, and shows that captions produced this way are good enough to fine-tune CLIP into a meme-text retriever that outperforms the original CLIP on human-annotated test sets. The headline numbers are Recall@1 for meme-caption retrieval rising from 0.680 to 0.770 on MemeCap and from 0.696 to 0.861 on CM50. The value, if the claim holds, is a scalable recipe: template knowledge plus a strong vision-language model produces training captions for meme retrieval without manual labeling.

What carries the argument

The machinery is CM50 plus the annotation pipeline that creates it: 33,172 memes drawn from 50 popular templates, each annotated by GPT-4o using a prompt that first asks the model to explain the meme and then outputs an image caption, meme caption, embedded text, literary-device label, and emotion label, guided by 50 expert-written template contexts from a meme knowledge base. The resulting meme captions are the only text used to fine-tune CLIP-ViT-L/14@336px with a cosine-annealed learning-rate schedule, gradient accumulation, and 20 epochs, producing the mtrCLIP retrieval model.

What would settle it

Run mtrCLIP on MemeCap instances whose templates are not among the 50 in CM50; if the Recall@1 gain over the original CLIP disappears or reverses on this subset, the improvement is template-specific rather than a general meme-understanding gain.

Watch

Extended reading notes

Core claim

The paper's central claim is that fine-tuning CLIP-ViT-L/14@336px solely on meme captions generated by GPT-4o for the CM50 dataset substantially improves meme-text retrieval over the original CLIP. On MemeCap, meme-caption retrieval Recall@1 rises from 0.680 to 0.770; on CM50 it rises from 0.696 to 0.861, with embedded-text retrieval also improving on both datasets. The fine-tuned model performs slightly worse than the original on image captions and meme titles, but the paper takes the meme-caption gains as evidence that a single cross-modal embedding model can capture meme semantics when given large-scale, template-grounded training captions.

Load-bearing premise

The retrieval gain rests on GPT-4o's meme captions being accurate and human-like enough that a model trained on them learns general meme-to-meaning alignment rather than template-specific phrasing.

Editorial extensions

If this is right

  • CM50 provides 33,172 memes across 50 templates, each with an image caption, meme caption, embedded text, and literary-device label, enabling large-scale meme studies.
  • The GPT-4o annotation pipeline with template context produces captions that human evaluators ranked above those from an open-source alternative and that reach a BLEURT of 0.525 on MemeCap, which the paper describes as human-level.
  • mtrCLIP, fine-tuned only on CM50 meme captions, raises meme-caption retrieval Recall@1 from 0.680 to 0.770 on MemeCap and from 0.696 to 0.861 on CM50.
  • The fine-tuned model also improves embedded-text retrieval on both datasets but performs slightly worse than the original CLIP on image captions and meme titles.
  • The annotation framework extends to new templates by supplying a knowledge-grounded prompt, making further dataset expansion possible without manual labeling.

Reading between the lines

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

  • If the transfer to MemeCap's human captions survives on templates outside the 50 in CM50, then synthetic captions could substitute for a large fraction of human meme annotation across other retrieval and classification tasks.
  • Because literary-device labeling plateaus at 0.39 macro F1, the captions may encode surface meaning while missing figurative structure; the retrieval gains could partly come from template-specific vocabulary rather than general meme semantics, an effect that ablating template identity would expose.
  • A natural stress test is to fine-tune on CM50 captions but evaluate on a held-out set of templates from another source, such as MemeCap templates not in CM50; a drop to baseline would bound the approach to known templates.
  • The slight regression on image captions and titles suggests mtrCLIP is specialized for meme-caption semantics and should not be treated as a general meme embedder without additional training data.
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 / 8 minor

Summary. The paper introduces CM50, a dataset of 33,172 memes from 50 popular templates, together with an automated annotation pipeline that uses GPT-4o with template context to produce image captions, meme captions, embedded text, and literary-device labels. It also presents mtrCLIP, a CLIP-ViT-L/14@336px model fine-tuned on CM50 meme captions, and reports retrieval results on MemeCap and CM50. The headline claim is that fine-tuning on the GPT-4o-generated CM50 meme captions transfers to human-annotated memes, improving MemeCap meme-caption retrieval at R@1 from 0.680 to 0.770. The paper also reports human-preference and automatic evaluations of the annotation pipeline, including literary-device macro F1 scores on Figmemes.

Significance. If the transfer result holds, CM50 and the annotation pipeline would provide a scalable route to meme annotation, and mtrCLIP would be a useful retrieval tool for meme analysis. The paper has clear strengths: it evaluates on external datasets (MemeCap, Figmemes), reports extensive prompt-engineering experiments, and makes code available. The central retrieval claim is, however, not yet established because the MemeCap improvement is not separated from template overlap with the 50 CM50 templates, and the annotation-quality claims are weakened by low literary-device F1 scores and a task-mismatched 'human-level' comparison.

major comments (3)
  1. [§4.2, Table 3] The claimed transfer to human-annotated memes is not separated from template overlap. CM50 contains exactly 50 templates, and the paper's own pipeline in §3.3.1 identifies 42 MemeCap test memes as instances of these templates. The MemeCap R@1 improvement from 0.680 to 0.770 corresponds to roughly 50 of 559 test items, so the gain could be concentrated among memes that share templates with CM50 and reflect memorized template layouts or caption styles rather than general meme semantics. The CM50 results in Table 4 are in-distribution by construction and cannot establish transfer. Please provide stratified results for template-overlapping versus non-overlapping MemeCap items, per-template R@K, and confidence intervals or bootstrapped significance tests. Without this, the headline 'fine-tuning improves transfer to human annotations' claim is not supported.
  2. [§3.3.2 and §5] The 'human-level annotation' claim is not supported by the evidence presented. The BLEURT score of 0.525 is compared to 0.448 from Bhavya et al. (2022), which is a different task (analogy generation) with a different evaluation protocol; this is not a valid human baseline for meme captioning. At the same time, the paper reports a macro F1 of only 0.39 for literary-device labeling on Figmemes and acknowledges that GPT-4o over-interprets and struggles with figurative language. The conclusion's statement that the method achieves 'close to human-level performance in captioning tasks' should be tempered or supported by a task-matched human annotation study on the same data and metrics.
  3. [§3.3.2, Tables 9–11] The literary-device evaluation is too weak to support the dataset's annotation-quality claims. The best macro F1 on the full Figmemes test set is 0.39, and on templatic subsets it ranges from 0.25 to 0.33 across prompt variants, which is substantially below the performance of models trained on Figmemes in the original paper. The paper explicitly notes that few-shot prompting did not help and that the three-step reasoning prompt performs on par with the baseline. Since one of the three stated contributions is an annotation framework producing literary-device labels, the manuscript should either weaken the claim that these labels are high-quality or demonstrate a downstream setting in which the labels are reliable enough for training or evaluation.
minor comments (8)
  1. [§3.1] The dataset size is reported inconsistently as 'over 33,000', '33,172', and '33,173' in different places; please unify the exact count.
  2. [§3.1] There is a typo 'ensamble' in 'using it for ensamble learning'; it should be 'ensemble'.
  3. [§3.2] The model name 'LlaV A' is used with inconsistent spacing and casing; it should be LLaVA consistently.
  4. [§3.3.1] The text refers to 'Figmeme' and 'Figmemes' inconsistently; also, '46 template-instance pairs in Figmeme' should read 'Figmemes'.
  5. [§4.2] There is an incomplete sentence: 'Between 5 and 20 epochs.' appears to be a fragment; please complete or remove it.
  6. [Table 10] Several cells in Table 10 are blank for the multi-reference metrics, making it difficult to compare conditions; please fill in all values or mark them explicitly as not applicable.
  7. [§2.2] The spelling 'MemeGaurd' appears in the text while the reference and dataset name are 'MemeGuard'; please correct for consistency.
  8. [Table 3 caption] The caption says 'R@K Avg. refers to the average R@K values' but then defines 'Mean' as the average of R@1, R@5, and R@10; please clarify the distinction between these two reported quantities.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: the main retrieval claim transfers from CM50 to the external MemeCap test set; only minor non-load-bearing self-citations appear.

full rationale

The central claim is that fine-tuning CLIP on CM50 meme captions improves meme-caption retrieval on MemeCap, a human-annotated external benchmark. This is not circular: mtrCLIP is trained on GPT-4o-generated captions from CM50 and tested on MemeCap's human captions, so the MemeCap R@1 improvement from 0.680 to 0.770 is a genuine out-of-distribution transfer measurement rather than an identity. The CM50 results in Table 4 are in-distribution and do not by themselves establish transfer, but they are reported alongside the external MemeCap results rather than substituted for them. The paper's annotation-quality evaluation uses Figmemes and MemeCap, external test sets with human labels, and the reported literary-device macro F1 of 0.39 is an honest weakness, not a circular validation. The only self-referential elements are the adoption of the TLC template-matching pipeline from KYMDB (Bates et al., 2023) and template metadata from KYMDB, where Bates et al. shares a co-author with the present paper. These citations are used to identify templatic memes and to supply template context, not to establish the retrieval improvement or the annotation quality; thus they are not load-bearing for the paper's central claims. No step in the derivation reduces by construction to a fitted parameter or to a self-citation chain, so the paper is not circular.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claims rest on the quality of GPT-4o annotations and on the choice of dataset filtering thresholds. The dataset is generated rather than human-annotated, and the full-scale annotation quality is not verified. The fine-tuning hyperparameters were selected on validation data, so they are free parameters with respect to the reported test results.

free parameters (6)
  • Minimum instances per template = 150
    Templates with fewer than 150 filtered memes were excluded (Section 3.1), a threshold chosen by hand to guarantee enough data per template.
  • Text length filter = unspecified
    Memes were kept only if they had 'text of sufficient length'; the exact length criterion is not reported (Section 3.1).
  • Title difference filter = title differs from base template
    Memes whose title equals the base template name were removed (Section 3.1).
  • LPIPS threshold = 1
    Used in template-instance matching to verify retrieved template pairs (Section 3.3.1).
  • CLIP matching thresholds = 30 and 1
    Thresholds for the concatenated embedding and fancy fusion embedding methods from KYMDB (Section 3.3.1).
  • Fine-tuning hyperparameters = peak LR 1e-5, batch size 2048/2400, weight decay 0.1, 20 epochs
    Selected via Ray Tune on a single A100 over three days; they are fitted to validation performance on CM50.
assumptions (4)
  • domain assumption GPT-4o-generated meme captions are of sufficient quality for CLIP fine-tuning to transfer to human-annotated captions.
    Section 3.3 reports automatic evaluation on MemeCap and Figmemes, but the full 33k set is not human-verified.
  • domain assumption The template context from KnowYourMeme (About sections) accurately describes the meme template and improves annotation.
    Template context is injected into the GPT-4o prompts (Section 3.2.1) and is claimed to help, particularly for smaller models, but the effect is only evaluated on small subsets.
  • domain assumption The ImgFlip filtering heuristics yield a representative distribution of templatic memes that supports generalization to MemeCap.
    The filters (at least 150 instances, long text, unique title) are described in Section 3.1 and may bias CM50 toward certain meme types; the authors do not analyze distribution shift.
  • domain assumption The retrieval evaluation on MemeCap is a valid measure of meme-text alignment in the CLIP embedding space.
    The paper assumes cosine similarity in the fine-tuned CLIP space is a meaningful retrieval metric, following standard practice from Section 4.1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Large Vision-Language Models for Knowledge-Grounded Data Annotation of Memes." pith.science (2026). https://pith.science/paper/PT7T23CL

@misc{pith2026250113851,
  author       = {Pith},
  title        = {Pith review of: Large Vision-Language Models for Knowledge-Grounded Data Annotation of Memes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PT7T23CL}},
  note         = {Machine review of arXiv:2501.13851}
}
read the original abstract

Memes have emerged as a powerful form of communication, integrating visual and textual elements to convey humor, satire, and cultural messages. Existing research has focused primarily on aspects such as emotion classification, meme generation, propagation, interpretation, figurative language, and sociolinguistics, but has often overlooked deeper meme comprehension and meme-text retrieval. To address these gaps, this study introduces ClassicMemes-50-templates (CM50), a large-scale dataset consisting of over 33,000 memes, centered around 50 popular meme templates. We also present an automated knowledge-grounded annotation pipeline leveraging large vision-language models to produce high-quality image captions, meme captions, and literary device labels overcoming the labor intensive demands of manual annotation. Additionally, we propose a meme-text retrieval CLIP model (mtrCLIP) that utilizes cross-modal embedding to enhance meme analysis, significantly improving retrieval performance. Our contributions include:(1) a novel dataset for large-scale meme study, (2) a scalable meme annotation framework, and (3) a fine-tuned CLIP for meme-text retrieval, all aimed at advancing the understanding and analysis of memes at scale.

Figures

Figures reproduced from arXiv: 2501.13851 by the authors.

Figure 1
Figure 1. Diagram of meme-text retrieval, illustrated [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Collage of memes found in CM50 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Example data annotation using our pipeline [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Left: Distribution of lengths (in tokens) for image captions, meme captions, and embedded text; Right: [PITH_FULL_IMAGE:figures/full_fig_p018_4.png]
Figure 5
Figure 5. Figure 5: CM50 dataset statistics of different templates present in our dataset. [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 23 canonical work pages

  1. [1]

    Luke Bates, Peter Ebert Christensen, Preslav Nakov, and Iryna Gurevych. 2023. A template is all you meme. arXiv preprint arXiv:2311.06649

  2. [2]

    Bhavya Bhavya, Jinjun Xiong, and ChengXiang Zhai. 2022. Analogy generation by prompting large language models: A case study of instructgpt. arXiv preprint arXiv:2210.04186

  3. [3]

    Branislav Buchel. 2012. Internet memes as means of communication. Brno: Masaryk University

  4. [4]

    Min Cao, Shiping Li, Juntao Li, Liqiang Nie, and Min Zhang. 2022. Image-text retrieval: A survey on recent research and development. arXiv preprint arXiv:2203.14713

  5. [5]

    Tuhin Chakrabarty, Debanjan Ghosh, Adam Poliak, and Smaranda Muresan. 2021 a . Figurative language in recognizing textual entailment. arXiv preprint arXiv:2106.01195

  6. [6]

    Tuhin Chakrabarty, Xurui Zhang, Smaranda Muresan, and Nanyun Peng. 2021 b . Mermaid: Metaphor generation with symbolism and discriminative decoding. arXiv preprint arXiv:2103.06779

  7. [7]

    Peter Ebert Christensen, Frederik Warburg, Menglin Jia, and Serge Belongie. 2022. https://arxiv.org/abs/2209.00495 Searching for structure in unfalsifiable claims . Preprint, arXiv:2209.00495

  8. [8]

    Peter Ebert Christensen, Srishti Yadav, and Serge Belongie. 2023. https://arxiv.org/abs/2309.10359 Prompt, condition, and generate: Classification of unsupported claims with in-context learning . Preprint, arXiv:2309.10359

Show all 42 references
  1. [9]

    it gets better

    Noam Gal, Limor Shifman, and Zohar Kampf. 2016. “it gets better”: Internet memes and the construction of collective identity. New media & society, 18(8):1698--1714

  2. [10]

    Biagio Grasso, Valerio La Gatta, Vincenzo Moscato, and Giancarlo Sperl \` . 2024. Kermit: Knowledge-empowered model in harmful meme detection. Information Fusion, 106:102269

  3. [11]

    facebook revolution

    Summer Harlow et al. 2013. It was a" facebook revolution": Exploring the meme-like spread of narratives during the egyptian protests. Revista de comunicaci \'o n , 12(1):59--82

  4. [12]

    Muzhaffar Hazman, Susan McKeever, and Josephine Griffith. 2024. https://arxiv.org/abs/2407.11861 What makes a meme a meme? identifying memes for memetics-aware dataset creation . Preprint, arXiv:2407.11861

  5. [13]

    Zeyu He, Chieh-Yang Huang, Chien-Kuang Cornelia Ding, Shaurya Rohatgi, and Ting-Hao Kenneth Huang. 2024. If in a crowdsourced data annotation pipeline, a gpt-4. In Proceedings of the CHI Conference on Human Factors in Computing Systems, pages 1--25

  6. [14]

    Cille Hvass Holm. 2021. https://doi.org/10.7146/lev.v0i7.125340 What do you meme? the sociolinguistic potential of internet memes . Leviathan: Interdisciplinary Journal in English, (7):1–20

  7. [15]

    EunJeong Hwang and Vered Shwartz. 2023. Memecap: A dataset for captioning and interpreting memes. arXiv preprint arXiv:2305.13703

  8. [16]

    Prince Jha, Raghav Jain, Konika Mandal, Aman Chadha, Sriparna Saha, and Pushpak Bhattacharyya. 2024. https://doi.org/10.18653/v1/2024.acl-long.439 M eme G uard: An LLM and VLM -based framework for advancing content moderation via meme intervention . In Proceedings of the 62nd ...

  9. [17]

    Chao Jia, Yinfei Yang, Ye Xia, Yi-Ting Chen, Zarana Parekh, Hieu Pham, Quoc Le, Yun-Hsuan Sung, Zhen Li, and Tom Duerig. 2021. Scaling up visual and vision-language representation learning with noisy text supervision. In International conference on machine learning, pages 4904...

  10. [18]

    Saurav Joshi, Filip Ilievski, and Luca Luceri. 2024. https://doi.org/10.1145/3589335.3651970 Contextualizing internet memes across social media platforms . In Companion Proceedings of the ACM Web Conference 2024, WWW '24, page 1831–1840, New York, NY, USA. Association for Comp...

  11. [19]

    Hyunjae Kim, Seunghyun Yoon, Trung Bui, Handong Zhao, Quan Tran, Franck Dernoncourt, and Jaewoo Kang. 2024. Fine-tuning clip text encoders with two-step paraphrasing. arXiv preprint arXiv:2402.15120

  12. [20]

    Chen Liu, Gregor Geigle, Robin Krebs, and Iryna Gurevych. 2022. Figmemes: A dataset for figurative language identification in politically-opinionated memes. In Proceedings of the 2022 conference on empirical methods in natural language processing, pages 7069--7086

  13. [21]

    Haotian Liu, Chunyuan Li, Yuheng Li, and Yong Jae Lee. 2024. Improved baselines with visual instruction tuning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 26296--26306

  14. [22]

    Yinhan Liu. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692, 364

  15. [23]

    Levente Murgás, Marcell Nagy, Kate Barnes, and Roland Molontay. 2024. https://arxiv.org/abs/2408.08126 Decoding memes: A comparative study of machine learning models for template identification . Preprint, arXiv:2408.08126

  16. [24]

    Shraman Pramanick, Aniket Roy, and Vishal M Patel. 2022. Multimodal learning using optimal transport for sarcasm and humor detection. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 3930--3940

  17. [25]

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. 2021. Learning transferable visual models from natural language supervision. In International conference on machine learni...

  18. [26]

    Hanoona Rasheed, Muhammad Uzair Khattak, Muhammad Maaz, Salman Khan, and Fahad Shahbaz Khan. 2023. Fine-tuned clip models are efficient video learners. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6545--6554

  19. [27]

    Arkadiy Saakyan, Shreyas Kulkarni, Tuhin Chakrabarty, and Smaranda Muresan. 2024. V-flute: Visual figurative language understanding with textual explanations. arXiv preprint arXiv:2405.01474

  20. [28]

    Chhavi Sharma, Deepesh Bhageria, William Scott, Srinivas Pykl, Amitava Das, Tanmoy Chakraborty, Viswanath Pulabaigari, and Bjorn Gamback. 2020. Semeval-2020 task 8: Memotion analysis--the visuo-lingual metaphor! arXiv preprint arXiv:2008.03781

  21. [29]

    Shivam Sharma, Siddhant Agarwal, Tharun Suresh, Preslav Nakov, Md Shad Akhtar, and Tanmoy Chakraborty. 2023. What do you meme? generating explanations for visual semantic role labelling in memes. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, page...

  22. [30]

    Will Styler. 2020. https://wstyler.ucsd.edu/talks/meme_linguistics.html#/ The linguistics of memes . https://wstyler.ucsd.edu/talks/meme_linguistics.html#/; accessed 26-June-2023

  23. [31]

    Zhen Tan, Dawei Li, Song Wang, Alimohammad Beigi, Bohan Jiang, Amrita Bhattacharjee, Mansooreh Karami, Jundong Li, Lu Cheng, and Huan Liu. 2024. Large language models for data annotation: A survey. arXiv preprint arXiv:2402.13446

  24. [32]

    Kohtaro Tanaka, Hiroaki Yamane, Yusuke Mori, Yusuke Mukuta, and Tatsuya Harada. 2022. Learning to evaluate humor in memes based on the incongruity theory. In Proceedings of the Second Workshop on When Creative AI Meets Conversational AI, pages 81--93

  25. [33]

    E Meltem Tolunay et al. 2018. Dank learning: Generating memes using deep neural networks. arXiv: 1806.04510

  26. [34]

    Riccardo Tommasini, Filip Ilievski, and Thilini Wijesiriwardene. 2023. https://doi.org/10.1007/978-3-031-33455-9_21 Imkg: The internet meme knowledge graph . In The Semantic Web: 20th International Conference, ESWC 2023, Hersonissos, Crete, Greece, May 28–June 1, 2023, Proceed...

  27. [35]

    Suryatej Reddy Vyalla and Vishaal Udandarao. 2020. Memeify: A large-scale meme generation system. In Proceedings of the 7th ACM IKDD CoDS and 25th COMAD, pages 307--311

  28. [36]

    Han Wang and Roy Ka-Wei Lee. 2024. Memecraft: Contextual and stance-driven multimodal meme generation. In Proceedings of the ACM on Web Conference 2024, pages 4642--4652

  29. [37]

    Yixuan Wei, Han Hu, Zhenda Xie, Ze Liu, Zheng Zhang, Yue Cao, Jianmin Bao, Dong Chen, and Baining Guo. 2023. Improving clip fine-tuning performance. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 5439--5449

  30. [38]

    Wentao Yu and Dorothea Kolossa. 2023. https://ceur-ws.org/Vol-3555/paper6.pdf wentaorub at memotion 3: Ensemble learning for multi-modal meme classification (short paper) . In DE-FACTIFY@AAAI

  31. [39]

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. 2018. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 586--595

  32. [40]

    Naitian Zhou, David Jurgens, and David Bamman. 2023. Social meme-ing: Measuring linguistic variation in memes. arXiv preprint arXiv:2311.09130

  33. [41]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...

  34. [42]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.