Pith. sign in

REVIEW 3 major objections 5 minor 38 references

A Self-Learning Multimodal Approach for Fake News Detection

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper claims a three-component fake news detector—contrastive image learning, Q-Former fusion into a Vicuna-7B language model, and a two-layer classifier—reaches 88.88% accuracy on a filtered Fakeddit split, beating eight prior…

desk verdict The accuracy gain is real but the claimed self-learning mechanism has no path into the classifier, so the paper's central contribution is unsupported. read the letter →

arxiv 2412.05843 v1 pith:KCXTEYBH submitted 2024-12-08 cs.CL cs.LG

Hao Chen , Hui Guo , Baochen Hu , Shu Hu , Jinrong Hu , Siwei Lyu , Xi Wu , Xin Wang This is my paper · ORCID
classification cs.CLcs.LG
keywords fakenewsdetectionmultimodalclassificationcontrastivelearninglargelanguagemodelsQ-FormerFakedditimage-textalignmentautomaticweightedloss
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 proposes a multimodal fake news classifier that combines contrastive learning on images with a large-language-model fusion stage, and claims it outperforms eight prior detectors on the Fakeddit benchmark. On a filtered Fakeddit split, the model reports accuracy of 88.88%, precision of 86.40%, recall of 85.40%, and F1-score of 85.90%, beating the best baseline by roughly four accuracy points. The authors argue that contrastive learning helps in settings with limited labeled image data, that the learnable-query alignment is what lets the language model use visual evidence, and that an automatically weighted loss keeps the contrastive and classification objectives balanced. If these claims hold, the method would be a new state of the art for paired image-text misinformation detection and a template for combining self-supervised visual learning with LLM reasoning.

What carries the argument

The load-bearing machinery is a three-stage pipeline. A contrastive learning module trains a ViT image encoder under InfoNCE loss with a momentum encoder update $y_t = m y_{t-1} + (1-m)x_t$; a multimodal fusion module uses Q-Former's learnable queries to align image features with text and feeds the combined prompt into the Vicuna-7B language model; and a two-layer MLP with GELU produces the true/false output. The two losses are merged by an automatic weighted loss $\mathcal{L} \approx \frac{1}{2\sigma_1^2}\mathcal{L}_1 + \frac{1}{\sigma_2^2}\mathcal{L}_2 + \log(1+\sigma_1) + \log(1+\sigma_2)$, with learnable uncertainties $\sigma_1,\sigma_2$. The fusion stage uses Q-Former features rather than the contrastive module's output, so the contrastive branch and the classifier are trained under a shared total loss rather than through a direct feature path.

What would settle it

Run the same experiment with the contrastive module removed while keeping all data augmentation: if accuracy and F1-score do not drop, the contrastive branch is not load-bearing. Tracing gradients or input connections from the classification loss back through the fusion module would also settle whether the contrastive head can affect the prediction at all.

Watch

Extended reading notes

Core claim

On its own terms, the central discovery is that a three-component model—a contrastive image feature learner, a Q-Former-based multimodal fusion into a pre-trained large language model, and a small classification head—can classify fake news from paired image-text posts at 88.88% accuracy on Fakeddit, with balanced precision (86.40%), recall (85.40%), and F1-score (85.90%). The paper's ablation attributes the gains to multimodal fusion over single-modality input, and to the addition of the contrastive and infuse modules, with accuracy rising from 87.16% to 88.21% to 88.88%. The authors further claim the approach is data-efficient because contrastive learning exploits unlabeled images and the language model requires only modest fine-tuning.

Load-bearing premise

The load-bearing premise is that the contrastive learning module's image features actually reach the final classifier; the paper says fusion uses Q-Former features instead of the contrastive module's features, so if the contrastive branch never feeds the classifier, the reported accuracy gains cannot be attributed to self-learning.

Editorial extensions

If this is right

  • A large multimodal LLM fine-tuned with 563k training pairs can reach 88.88% accuracy on a filtered Fakeddit split, suggesting that model scale and prompt-based fusion are decisive for this task.
  • Multimodal fusion contributes about 5 to 7 accuracy points over text-only or image-only inputs on this benchmark (88.88% vs. 83.18% and 79.68%).
  • The ablation sequence (LLM only, then plus contrastive, then plus infuse) implies each added module improves accuracy, from 87.16% to 88.21% to 88.88%.
  • Automatically weighting the contrastive and classification losses avoids manual tuning and is presented as a way to fine-tune the large language model jointly with the image encoder.
  • Because the contrastive module is trained without labels, the same architecture could transfer to domains where labeled fake news examples are scarce.

Reading between the lines

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

  • A direct test of the paper's self-learning claim would be to route the contrastive encoder's features into the Q-Former alongside the frozen image encoder and re-measure accuracy; if accuracy rises, the contrastive branch is load-bearing, and if not, the gains likely come from the language model backbone and prompts.
  • The reported results are on a filtered Fakeddit split with entries lacking valid images removed, so an independent check would re-run every baseline on this exact filtered split before comparing with published numbers from the unfiltered benchmark.
  • The method's randomized use of four paraphrased questions as prompts is a general regularization idea for vision-language classification, though the paper does not ablate the number or wording of prompts.
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 / 5 minor

Summary. The paper proposes a multimodal fake-news detector for image-text pairs, combining a contrastive-learning image encoder with a Q-Former-based fusion module, a Vicuna-7B LLM, and a two-layer MLP classifier. The training loss is an uncertainty-weighted sum of InfoNCE and classification losses. Experiments on a filtered subset of Fakeddit report accuracy 88.88%, precision 86.40%, recall 85.40%, and F1-score 85.90%, which the authors claim outperforms EANN, CAFE, SpotFake, SpotFake+, MVAE, HMCAN, and VERITE. The paper also includes single-modality and module-level ablations.

Significance. The problem is important, and the use of a public benchmark and pre-trained components are strengths. The reported gains over the listed baselines are large, but the comparison is confounded by model scale and by the filtered, nonstandard benchmark. The central claim that the self-supervised contrastive module materially contributes to the final prediction is not currently supported: the fusion equations do not show a path from the contrastive module into the classifier, and the full-model F1 is lower than one of the ablated configurations. If these points are resolved, the paper would be a useful contribution to LLM-based multimodal fake-news detection; as written, the evidence is suggestive but not conclusive.

major comments (3)
  1. [III.B, Eqs. (3)-(4)] The classification input is defined as eprompt = etext + f_Q-Former(eimg) and E = LLM(eprompt), with no term originating from the contrastive module described in Section III.A. The text explicitly states, 'Instead of using the image features from contrastive learning module, we would like to leverage Q-Former...', and Figure 2 shows the contrastive module as a separate component. If the Q-Former's frozen image encoder is not the ViT trained with the InfoNCE loss, then the contrastive module cannot influence the final prediction, contradicting Contribution 1 and the paper's title. If the Q-Former uses the contrastively pretrained ViT as its frozen encoder, that coupling must be stated explicitly; currently the architecture is ambiguous on this load-bearing point.
  2. [Table IV] The full model (Experiment C) reports F1 85.90, which is lower than Experiment B's 86.02, and recall drops from 91.45 to 85.40 when the infuse module is added. Only accuracy improves from 88.21 to 88.88. The claim in Section V.A that 'Ours' achieves the highest F1 among all configurations is therefore not supported by the ablation, and the conclusion that the learnable alignment module significantly improves the model is at best metric-dependent. This needs to be discussed and reconciled.
  3. [IV.A, IV.C, Table II] The state-of-the-art comparison is on a filtered subset of Fakeddit from which entries without images were removed, with no evaluation on the unfiltered benchmark or comparison to previously reported Fakeddit numbers. The baselines are all substantially smaller models; the paper itself notes model scale as a factor in Section V.A. Moreover, no error bars, multiple seeds, or statistical significance tests are reported. The claim of superiority over 'state-of-the-art' is therefore not yet established; a same-scale LLM-based baseline and variance estimates are needed.
minor comments (5)
  1. [References [15], [30], [31]] Reference [15] does not describe the CAFE fake-news model, and references [30] and [31] are the same SpotFake citation; please correct and deduplicate.
  2. [IV.B] The manuscript omits the optimizer, learning rate, scheduler, and fine-tuning details for Vicuna, and the reported feature dimension 'batch size × 12 × 1048' is unexplained; please provide full reproducibility details.
  3. [III.D] The four prompts are randomly selected during training; report whether the selection is per sample or per batch/epoch and add a prompt-sensitivity analysis, since the prompts are a hand-written design choice.
  4. [Throughout] Typographical errors such as 'LLMs are excel at this task', 'authentic of news', 'the the', 'Monument configuration' in Fig. 3, and 'approapriate' should be corrected.
  5. [Eq. (6)] The second term appears to be written as 1/σ2^2 times L2 rather than 1/(2σ2^2) times L2; if this is intentional, explain the deviation from the AWL formulation in [17].

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the reported gains are measured against an external benchmark, and no equation or fitted parameter reduces the central claim to its own inputs.

full rationale

The paper's central claim is that a contrastive-learning module combined with a MiniGPT-4-style fusion and classifier achieves 88.88% accuracy on Fakeddit. The evaluation is against a public external benchmark, with performance measured on a held-out test split; the results are not recovered from training labels, fitted parameters, or the paper's own assumptions. The architecture equations (3)-(5) define a feed-forward pipeline (eprompt = etext + f_Q-Former(eimg), E = LLM(eprompt), L2 = MLP(E)), and none of these definitions embeds the output label or the evaluation metric, so there is no self-definitional reduction. The learnable sigma weights in Equation (6) are standard uncertainty-based multi-task weighting and do not encode the classification outcome. Self-citations in the paper are confined to background references on deepfake detection and do not carry the load of the main result. One can question whether the contrastive module actually contributes to the reported performance, since Section III.B states that the fusion uses Q-Former features 'instead of using the image features from contrastive learning module,' but that is an architectural-support or attribution problem, not circular reasoning; the model is not defined in terms of its own predictions, and the reported gains remain externally falsifiable. No circularity is therefore present.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the transferability of pretrained components (Q-Former, Vicuna) and on a filtered Fakeddit benchmark. The learned AWL weights and the hand-set temperature and momentum are the main free parameters; no entity is invented. The most fragile assumptions are that the contrastive module is actually connected to the classifier and that the filtered dataset preserves comparability.

free parameters (4)
  • sigma_1 = learned during training
    Uncertainty weight for contrastive loss L1 in AWL (Eq. 6), learned during training; affects the balance of losses.
  • sigma_2 = learned during training
    Uncertainty weight for classification loss L2 in AWL (Eq. 6), learned during training.
  • temperature tau = not reported
    Temperature in InfoNCE loss (Eq. 2), chosen by hand; controls contrastive sharpness.
  • momentum m = not reported
    Momentum coefficient in Eq. 1 for the momentum encoder; chosen by hand.
assumptions (3)
  • domain assumption Q-Former alignment between a frozen image encoder and Vicuna yields features that support fake-news classification.
    The method relies on BLIP-2/MiniGPT-4's trained Q-Former and Vicuna, assumed transferable to misinformation detection (Sec III.B).
  • domain assumption Filtering Fakeddit to samples with valid images preserves the benchmark's difficulty and comparability.
    The paper removes entries without images and reports results only on the filtered set (Sec IV-A), without comparing to the original Fakeddit split.
  • ad hoc to paper The four hand-written prompts elicit the LLM's judgment without introducing bias.
    Prompts in Sec III.D are randomly chosen and are paper-specific; their effect on classification is not analyzed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Self-Learning Multimodal Approach for Fake News Detection." pith.science (2026). https://pith.science/paper/KCXTEYBH

@misc{pith2026241205843,
  author       = {Pith},
  title        = {Pith review of: A Self-Learning Multimodal Approach for Fake News Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KCXTEYBH}},
  note         = {Machine review of arXiv:2412.05843}
}
read the original abstract

The rapid growth of social media has resulted in an explosion of online news content, leading to a significant increase in the spread of misleading or false information. While machine learning techniques have been widely applied to detect fake news, the scarcity of labeled datasets remains a critical challenge. Misinformation frequently appears as paired text and images, where a news article or headline is accompanied by a related visuals. In this paper, we introduce a self-learning multimodal model for fake news classification. The model leverages contrastive learning, a robust method for feature extraction that operates without requiring labeled data, and integrates the strengths of Large Language Models (LLMs) to jointly analyze both text and image features. LLMs are excel at this task due to their ability to process diverse linguistic data drawn from extensive training corpora. Our experimental results on a public dataset demonstrate that the proposed model outperforms several state-of-the-art classification approaches, achieving over 85% accuracy, precision, recall, and F1-score. These findings highlight the model's effectiveness in tackling the challenges of multimodal fake news detection.

Figures

Figures reproduced from arXiv: 2412.05843 by the authors.

Figure 1
Figure 1. An example of fake news (mismatching image-text) from dataset [ [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall structure of multimodal fake news detection. The model is composed of three components, contrastive learning module is for learning [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Monument configuration for contrastive learning [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Q-Former structure adopted from [18] that are designed to interact with the input data in a way that helps extract relevant information and establish meaningful connections. The structure is illustrated in [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 31 canonical work pages

  1. [15]

    Cafe: Catastrophic data leakage in vertical federated learning

    Xiao Jin, Pin-Yu Chen, Chia-Yi Hsu, Chia-Mu Yu, and Tianyi Chen. Cafe: Catastrophic data leakage in vertical federated learning. In M. Ranzato, A. Beygelzimer, Y . Dauphin, P.S. Liang, and J. Wortman Vaughan, editors, Advances in Neural Information Processing Systems , volume 34, pages 994–1006. Curran Associates, Inc., 2021. 6

    work page 2021

  2. [1]

    Riegler, P

    Shivangi Aneja, Cise Midoglu, Duc-Tien Dang-Nguyen, Sohail Ahmed Khan, M. Riegler, P. Halvorsen, Christoph Bregler, and Balu Adsumilli. Acm multimedia grand challenge on detecting cheapfakes. ArXiv, abs/2207.14534, 2022. 2

    arXiv 2022

  3. [2]

    Bagozzi, Rajni Goel, Brunilda Lugo-De-Fabritz, Kelly Knickmeier-Cummings, and Karthik Balasubramanian

    Benjamin E. Bagozzi, Rajni Goel, Brunilda Lugo-De-Fabritz, Kelly Knickmeier-Cummings, and Karthik Balasubramanian. A framework for enhancing social media misinformation detection with topical-tactics. Digital Threats: Research and Practice , 2024. 1, 2

    work page 2024

  4. [3]

    A survey on fake news and rumour detection techniques

    Alessandro Bondielli and Francesco Marcelloni. A survey on fake news and rumour detection techniques. Information Sciences , 497:38–55,

    work page

  5. [4]

    Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-V oss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M. Ziegler, Jeff Wu, Clemens Winter, Christopher Hesse, Mark Chen, Eric Sigler, Ma teusz Litwin,...

    arXiv 2005

  6. [5]

    Deepfake: An overview

    Anupama Chadha, Vaibhav Kumar, Sonu Kashyap, and Mayank Gupta. Deepfake: An overview. 2021. 1

    work page 2021

  7. [6]

    Vicuna: An open-source chatbot impressing gpt-4 with 90%* chatgpt quality, march 2023

    Wei-Lin Chiang, Zhuohan Li, Zi Lin, Ying Sheng, Zhanghao Wu, Hao Zhang, Lianmin Zheng, Siyuan Zhuang, Yonghao Zhuang, Joseph E Gonzalez, et al. Vicuna: An open-source chatbot impressing gpt-4 with 90%* chatgpt quality, march 2023. URL https://lmsys. org/blog/2023- 03-30-vicuna, 3(5), 2023. 4

    work page 2023

  8. [7]

    Bert: Pre-training of deep bidirectional transformers for language under- standing

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for language under- standing. In North American Chapter of the Association for Computa- tional Linguistics, 2019. 2

    work page 2019

Show all 38 references
  1. [8]

    Multimodal multi-image fake news detection

    Anastasia Giachanou, Guobiao Zhang, and Paolo Rosso. Multimodal multi-image fake news detection. In 2020 IEEE 7th International Conference on Data Science and Advanced Analytics (DSAA) , pages 647–654, 2020. 2

    2020

  2. [9]

    Fake news detection using naive bayes classifier

    Mykhailo Granik and V olodymyr Mesyura. Fake news detection using naive bayes classifier. In 2017 IEEE First Ukraine Conference on Electrical and Computer Engineering (UKRCON) , pages 900–903,

    2017

  3. [10]

    The future of misinformation detection: New perspectives and trends

    Bin Guo, Yasan Ding, Lina Yao, Yunji Liang, and Zhiwen Yu. The future of misinformation detection: New perspectives and trends. ArXiv, abs/1909.03654, 2019. 2

    1909 arXiv

  4. [11]

    Eyes tell all: Irregular pupil shapes reveal gan-generated faces

    Hui Guo, Shu Hu, Xin Wang, Ming-Ching Chang, and Siwei Lyu. Eyes tell all: Irregular pupil shapes reveal gan-generated faces. In ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages 2904–2908. IEEE, 2022. 1

    2022

  5. [12]

    Open- eye: An open platform to study human performance on identifying ai- synthesized faces

    Hui Guo, Shu Hu, Xin Wang, Ming-Ching Chang, and Siwei Lyu. Open- eye: An open platform to study human performance on identifying ai- synthesized faces. arXiv preprint arXiv:2205.06680 , 2022. 2

    2022 arXiv

  6. [13]

    Robust attentive deep neural network for exposing gan-generated faces

    Hui Guo, Shu Hu, Xin Wang, Ming-Ching Chang, and Siwei Lyu. Robust attentive deep neural network for exposing gan-generated faces. IEEE Access, 2022. 2

    2022

  7. [14]

    Fake news detection using deep recurrent neural networks

    Tao Jiang, Jian Ping Li, Amin Ul Haq, and Abdus Saboor. Fake news detection using deep recurrent neural networks. In 2020 17th International Computer Conference on Wavelet Active Media Technology and Information Processing (ICCWAMTIP), pages 205–208, 2020. 1, 2

    2020

  8. [16]

    Fndnet – a deep convolutional neural network for fake news detection

    Rohit Kumar Kaliyar, Anurag Goswami, Pratik Narang, and Soumendu Sinha. Fndnet – a deep convolutional neural network for fake news detection. Cognitive Systems Research , 61:32–44, 2020. 1, 2

    2020

  9. [17]

    Multi-task learning using uncertainty to weigh losses for scene geometry and semantics

    Alex Kendall, Yarin Gal, and Roberto Cipolla. Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7482–7491, 2018. 5

    2018

  10. [18]

    Blip-2: bootstrapping language-image pre-training with frozen image encoders and large language models

    Junnan Li, Dongxu Li, Silvio Savarese, and Steven Hoi. Blip-2: bootstrapping language-image pre-training with frozen image encoders and large language models. In Proceedings of the 40th International Conference on Machine Learning , ICML’23. JMLR.org, 2023. 4

    2023

  11. [19]

    Malthouse

    Xinyi Li, Yongfeng Zhang, and Edward C. Malthouse. Large language model agent for fake news detection. ArXiv, abs/2405.01593, 2024. 2

    2024 arXiv

  12. [20]

    Fka-owl: Ad- vancing multimodal fake news detection through knowledge-augmented lvlms

    Xuannan Liu, Peipei Li, Huaibo Huang, Zekun Li, Xing Cui, Jiahao Liang, Lixiong Qin, Weihong Deng, and Zhaofeng He. Fka-owl: Ad- vancing multimodal fake news detection through knowledge-augmented lvlms. In ACM Multimedia, 2024. 2

    2024

  13. [21]

    Fake news, rumor, information pollution in social media and web: A contemporary survey of state-of-the-arts, challenges and opportunities

    Priyanka Meel and Dinesh Kumar Vishwakarma. Fake news, rumor, information pollution in social media and web: A contemporary survey of state-of-the-arts, challenges and opportunities. Expert Syst. Appl. , 153:112986, 2020. 2

    2020

  14. [22]

    Fakeddit: A new multimodal benchmark dataset for fine-grained fake news detection

    Kai Nakamura, Sharon Levy, and William Yang Wang. Fakeddit: A new multimodal benchmark dataset for fine-grained fake news detection. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 6149–6157, Marseille, France, May 2020. European Language Reso...

    2020

  15. [23]

    Fake news detection: A hybrid cnn-rnn based deep learning approach

    Jamal Abdul Nasir, Osama Subhani Khan, and Iraklis Varlamis. Fake news detection: A hybrid cnn-rnn based deep learning approach. Interna- tional Journal of Information Management Data Insights , 1(1):100007,

  16. [24]

    Petrantonakis

    Stefanos-Iordanis Papadopoulos, Christos Koutlis, Symeon Papadopou- los, and Panagiotis C. Petrantonakis. Verite: a robust benchmark for multimodal misinformation detection accounting for unimodal bias. Int. J. Multim. Inf. Retr. , 13:4, 2023. 6

    2023

  17. [25]

    Learning a deep dual-level network for robust deepfake detection

    Wenbo Pu, Jing Hu, Xin Wang, Yuezun Li, Shu Hu, Bin Zhu, Rui Song, Qi Song, Xi Wu, and Siwei Lyu. Learning a deep dual-level network for robust deepfake detection. Pattern Recognition, 130:108832, 2022. 1

    2022

  18. [26]

    Exploiting multi-domain visual information for fake news detection

    Peng Qi, Juan Cao, Tianyun Yang, Junbo Guo, and Jintao Li. Exploiting multi-domain visual information for fake news detection. In 2019 IEEE International Conference on Data Mining (ICDM) , pages 518– 527, 2019. 6

    2019

  19. [27]

    Hierarchical multi-modal contextual attention network for fake news detection

    Shengsheng Qian, Jinguang Wang, Jun Hu, Quan Fang, and Changsheng Xu. Hierarchical multi-modal contextual attention network for fake news detection. Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval , 2021. 6

    2021

  20. [28]

    Very deep convolutional networks for large-scale image recognition

    Karen Simonyan and Andrew Zisserman. Very deep convolutional networks for large-scale image recognition. CoRR, abs/1409.1556, 2014. 2

    2014 arXiv

  21. [29]

    Spotfake: A multi-modal framework for fake news detection

    Shivangi Singhal, Rajiv Ratn Shah, Tanmoy Chakraborty, Ponnurangam Kumaraguru, and Shin’ichi Satoh. Spotfake: A multi-modal framework for fake news detection. In 2019 IEEE Fifth International Conference on Multimedia Big Data (BigMM) , pages 39–47, 2019. 2

    2019

  22. [30]

    Spotfake: A multi-modal framework for fake news detection

    Shivangi Singhal, Rajiv Ratn Shah, Tanmoy Chakraborty, Ponnurangam Kumaraguru, and Shin’ichi Satoh. Spotfake: A multi-modal framework for fake news detection. In 2019 IEEE Fifth International Conference on Multimedia Big Data (BigMM) , pages 39–47, 2019. 6

    2019

  23. [31]

    Spotfake: A multi-modal framework for fake news detection

    Shivangi Singhal, Rajiv Ratn Shah, Tanmoy Chakraborty, Ponnurangam Kumaraguru, and Shin’ichi Satoh. Spotfake: A multi-modal framework for fake news detection. 2019 IEEE Fifth International Conference on Multimedia Big Data (BigMM) , pages 39–47, 2019. 6

    2019

  24. [32]

    Adapting fake news detection to the era of large language models

    Jinyan Su, Claire Cardie, and Preslav Nakov. Adapting fake news detection to the era of large language models. ArXiv, abs/2311.04917,

    arXiv

  25. [33]

    Sudhakar and K

    M. Sudhakar and K. P. Kaliyamurthie. Fake News Detection Approach Based on Logistic Regression in Machine Learning , pages 55–60. Springer Nature Singapore, Singapore, 2023. 2

    2023

  26. [34]

    Ting Wei Teo, Hui Na Chua, Muhammed Basheer Jasser, and Richard T.K. Wong. Integrating large language models and machine learning for fake news detection. In 2024 20th IEEE International Colloquium on Signal Processing & Its Applications (CSPA), pages 102– 107, 2024. 2

    2024

  27. [35]

    Gan- generated faces detection: A survey and new perspectives

    Xin Wang, Hui Guo, Shu Hu, Ming-Ching Chang, and Siwei Lyu. Gan- generated faces detection: A survey and new perspectives. ECAI, 2023. 1

    2023

  28. [36]

    Eann: Event adversarial neural networks for multi-modal fake news detection

    Yaqing Wang, Fenglong Ma, Zhiwei Jin, Ye Yuan, Guangxu Xun, Kishlay Jha, Lu Su, and Jing Gao. Eann: Event adversarial neural networks for multi-modal fake news detection. Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2018. 6

    2018

  29. [37]

    Minigpt-4: Enhancing vision-language understanding with advanced large language models

    Deyao Zhu, Jun Chen, Xiaoqian Shen, Xiang Li, and Mohamed El- hoseiny. Minigpt-4: Enhancing vision-language understanding with advanced large language models. ArXiv, abs/2304.10592, 2023. 2

    2023 arXiv

  30. [38]

    A robustly optimized BERT pre-training approach with post-training

    Liu Zhuang, Lin Wayne, Shi Ya, and Zhao Jun. A robustly optimized BERT pre-training approach with post-training. In Sheng Li, Maosong Sun, Yang Liu, Hua Wu, Kang Liu, Wanxiang Che, Shizhu He, and Gaoqi Rao, editors, Proceedings of the 20th Chinese National Confer- ence on Comp...

    2021

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.