Pith. sign in

REVIEW 4 major objections 6 minor 49 references

Multi-Label Contrastive Learning : A Comprehensive Study

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read This paper proposes a regularized multi-label contrastive loss, LREG, that cancels a repulsive gradient component from close positive pairs, and reports consistent Macro-F1 gains over existing losses across vision and text benchmarks.

desk verdict Useful empirical study with a plausible new loss, but the gradient derivation in Eqs. (8)-(9) is wrong and the shared-minimum claim is unsupported; worth reviewing but needs major revision. read the letter →

arxiv 2412.00101 v2 pith:2AQK7W4P submitted 2024-11-27 cs.LG

classification cs.LG
keywords multi-labelclassificationcontrastivelearningsupervisedlossgradientregularizationMacro-F1representationlow-dataregimelabelinteractions
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to establish that contrastive losses are a strong alternative to standard losses for multi-label classification when the label set is large, and that a specific gradient pathology limits them. It reports that existing supervised contrastive losses underperform on few-label datasets and ranking metrics, but excel on Macro-F1 when labels are numerous and interactions are dense. To fix the pathology, the authors derive the gradient of the multi-label contrastive loss and propose LREG, a regularized contrastive loss that cancels the contribution of positive pairs whose gradient direction mimics that of negative pairs. Empirically, LREG improves over prior contrastive losses on COCO, NUS-WIDE, AAPD, RCV1 and BGC, with the largest margins on Macro-F1 and in low-data regimes. The paper also argues that contrastive learning succeeds partly because of its optimization scheme, not only because it models label interactions.

What carries the argument

The central object is the regularized multi-label supervised contrastive loss $L_{\mathrm{REG}} = \frac{1}{|B|} \sum_{z_i \in B} \hat{\ell}(z_i)$, where each $\hat{\ell}$ is the weighted SupCon-style log-softmax over positive pairs plus the regularization $\ell_{\mathrm{reg}}(z_i) = -\sum_{j \in P(i)} \max(0, -\Lambda^i_j + \sigma_{j,i}.\mathrm{detach()}) \, z_j \cdot z_i$. Here $\Lambda$ is the normalized positive-pair weight and $\sigma$ is the softmax score; the max gate activates only when a positive pair's gradient would otherwise point like a negative pair's. The mechanism doing the work is gradient surgery: it removes one component of the positive-pair gradient without changing the location of the loss minima, so any gains come from the optimization trajectory rather than from a different objective.

What would settle it

Take Eq. (6) for a batch of one anchor, one positive, and one negative, and compute $\partial \ell / \partial z_k$ numerically with automatic differentiation. Compare the resulting vector to $z_k$ and to $z_i$. If it is proportional to $z_i$ in every positive case, the sign condition that activates $\ell_{\mathrm{reg}}$ (Eq. 11) would not follow from Eqs. (8)-(9) as written.

Watch

Extended reading notes

Core claim

The central claim is that in multi-label contrastive learning, once a positive pair becomes very close, the softmax score term $\sigma_{k,i}$ can exceed the normalized positive weight $\Lambda^i_k$, and then the gradient contribution of that positive pair points along the paired representation in the same direction as a negative pair's contribution, pushing the pair apart. On this reading, attraction and repulsion are asymmetric and the loss can contain an unwanted repulsive component. The paper's LREG loss adds a $\max(0, -\Lambda + \sigma) \cdot z_j \cdot z_i$ regularization with a detached $\sigma$, so that the offending component is removed while the global optimum of the contrastive loss is unchanged. The authors report that LREG matches or beats BCE, asymmetric loss, ZLPR and previous contrastive losses across six datasets, and that it improves representation alignment and Macro-F1, especially with 20% of the training data.

Load-bearing premise

The load-bearing assumption is that, in the log-softmax contrastive loss, the gradient of a positive pair is directed along the paired representation $z_k$; if standard differentiation with the anchor held fixed instead gives a direction along the anchor $z_i$, the claimed repulsion mechanism needs to be re-derived.

Editorial extensions

If this is right

  • If LREG's characterization is right, contrastive losses can be used as drop-in representation learners for multi-label text and image data, particularly when the label set is large.
  • Macro-F1 gains on large-label datasets should transfer to tasks where rare labels dominate the evaluation, since the regularization stabilizes close positive pairs.
  • Because LREG and its ablation share the same minima, the reported improvement is attributable to the optimization trajectory, supporting methods that focus on gradient dynamics rather than only objective design.
  • The regularization can be attached to the standard SupCon loss without hurting multi-class accuracy, so it appears safe to apply in the multi-label setting while leaving single-label performance unchanged.

Reading between the lines

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

  • The paper's gradient mechanism is stated for derivatives with respect to the paired representation $z_k$; under the usual chain rule for the same log-softmax with a fixed anchor $z_i$, the derivative is proportional to $z_i$. The empirical gains would still stand, but the repulsion narrative would need re-expression under that convention.
  • The strong results of prototype-only contrastive loss suggest that part of the gap between contrastive and BCE-based methods may come from an optimizer-friendly landscape rather than from label correlations; ablating label interactions while keeping the gradient gate would isolate that factor.
  • The reported mAP gap between ZLPR and all contrastive losses points to an untested complement: combining LREG with a ranking-oriented loss could yield a single objective that is strong on both clustering and ranking metrics.
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

4 major / 6 minor

Summary. The manuscript presents a broad empirical study of contrastive losses for multi-label classification in computer vision and NLP, across full-data and low-data regimes. It identifies a supposedly undesired gradient component in multi-label supervised contrastive loss and proposes a new loss, LREG, that adds a regularizer (Eq. 11) intended to cancel that component. The paper claims that LREG improves over prior contrastive and non-contrastive losses, particularly in Macro-F1 on datasets with many labels, and reports results on PASCAL, MS-COCO, NUS-WIDE, AAPD, RCV1, and BGC, with code released publicly.

Significance. If the theoretical story were correct, the paper would provide a useful systematic comparison of multi-label contrastive losses and a simple, practical regularizer with a clear geometric interpretation, backed by open-source code and a relatively broad set of benchmarks. The empirical finding that a gradient regularizer can improve Macro-F1 in large-label multi-label settings is potentially valuable. However, the central theoretical derivation in Sec. 3.4 is mathematically incorrect, and the claim in Sec. 4.2 that the regularized and original losses share the same minimum is not established. Since the proposed method's motivation and the paper's main narrative rest on these points, the contribution is currently conditional on correcting and re-validating that theory.

major comments (4)
  1. [Sec. 3.4, Eqs. (8) and (9)] The derivatives in Eqs. (8) and (9) are not the correct derivatives of Eq. (6). For the loss ℓ(z_i) = -Σ_{j∈P(i)} Λ_j log p_j with p_j = exp(z_j·z_i)/Σ_{k∈A(i)} exp(z_k·z_i), direct differentiation gives ∂ℓ/∂z_k = (σ_{k,i} - Λ_k) z_i for a positive k and ∂ℓ/∂z_k = σ_{k,i} z_i for a negative k. The printed expressions instead have these gradients proportional to z_k. Consequently, the geometric statement in Sec. 4.1 that a close positive pair's gradient "moves in the same direction as its representation z_k" is incorrect. The sign condition (σ - Λ) > 0 is still the right condition for the positive-pair gradient to point along +z_i, which under gradient descent repels z_k from z_i, so the regularizer in Eq. (11) can still cancel the repulsive component when implemented with autodiff; but the paper's printed derivation and explanation are not valid as they stand and must be corrected.
  2. [Sec. 4.2, Eq. (13)] Eq. (13) repeats the same direction error made in Eqs. (8) and (9): the gradient of ℓ_reg(z_i) with respect to a positive representation z_k is -max(0, -Λ_k+σ_{k,i}) z_i, not the printed expression proportional to z_k. Since the regularizer is actually evaluated through automatic differentiation in Algorithm 2, the implementation may well compute the correct cancellation, but the manuscript's mathematical description is inconsistent with its own algorithm. This needs to be fixed in both the main text and the algorithm description.
  3. [Sec. 4.2, claim about shared minima] The statement that ℂ(z_i) and ℂ(z_i)+ℓ_reg(z_i) "share the same minimum, as established in Eq. (10)" is not supported. Eq. (10) characterizes stationary points of the original loss ℓ, not of the regularized objective. In the region where σ_{k,i} > Λ_k, the regularizer exactly cancels the positive-pair gradient for that k, so the combined objective can have additional stationary points, plateaus, or modified basins. At minimum, the paper should prove that no new minima are introduced, or it should soften the claim and provide empirical evidence that the observed gains arise specifically from optimization dynamics rather than from altered stationary points.
  4. [Sec. 6.1.2 and Contributions] The claims that LREG "outperforms previous methods in almost all metrics" and "establishes state-of-the-art performance" are stronger than Table 2 supports. On PASCAL, LASY achieves higher Micro-F1 and Macro-F1 than LREG; on RCV1 and BGC, LREG's Micro-F1 and Hamming loss are worse than those of LZLPR. The more defensible reading of the tables is that LREG frequently improves Macro-F1 on datasets with many labels, while its advantages are mixed on Micro-F1 and Hamming. The abstract and Section 6 should be reworded to state this more precise conclusion.
minor comments (6)
  1. [Sec. 5.3 vs. Appendix A.2] The temperature hyperparameter is reported inconsistently: Sec. 5.3 says τ is set to 0.1, while Appendix A.2 states that "in all our previous experiments on both vision and language datasets, the temperature parameter was set to 0.01." Please clarify which value was actually used, since temperature strongly affects contrastive losses and the reported PRR values.
  2. [Eq. (16) vs. Eq. (18)] The paper defines ℂ(z_i) in Eq. (16) without the label-overlap reweighting f, then introduces a reweighted version in Eq. (18) with a hyperparameter α. The main results refer to "LREG (Eq. 15)" without specifying whether the final loss includes Eq. (18). Since Table 6 reports α values, presumably the experiments use Eq. (18), but this should be stated explicitly.
  3. [Table 3] Several entries in Table 3 appear to be malformed: "26.13." in the RCV1 row, "9.865" in the BGC Hamming column, and "86.57 3.099 97.77" for the LREG row on RCV1, which should probably read "73.09" for Macro-F1. Please correct these transcription errors.
  4. [Appendix A.6] The heading "Rankinp Metrics" contains a typo; it should be "Ranking Metrics." In addition, the mAP results are relegated to the appendix even though they directly support the paper's claim that contrastive losses are weaker on ranking-based metrics; consider mentioning the mAP conclusion in the main text.
  5. [Appendix A.5] The reference to "PASCAL VOC 2007 dataset [39]" is incorrect: [39] is the MS-COCO paper. PASCAL is [38]. Please fix the reference.
  6. [General] There are many missing spaces in mathematical expressions, e.g., "LP roto", "LM ulSupCon", and "LREG(Eq. 15)". A careful proofread would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: LREG is constructed from the paper's own gradient analysis and then evaluated on held-out test sets; self-citations are not load-bearing.

full rationale

The central claims are empirical: LREG is compared against baselines on fixed test splits, and its gains are not derived from fitted constants. The regularizer in Eq. (11) is engineered to cancel a gradient component identified in the paper's own Eqs. (8)-(9), so the cancellation is by construction, but that is the proposed method, not a prediction. The only self-citations ([18], [36]) are used as baselines or as design motivation for prototypes and label-frequency weighting; the LREG vs. LW/O-REG ablation provides independent evidence for the regularizer's effect. A separate mathematical concern is that differentiating Eq. (6) with respect to z_k gives a term proportional to z_i, not z_k, so Eqs. (8)-(9) and (13) may misstate the gradient direction; however, an erroneous derivation is a correctness risk, not circularity. Likewise, Eq. (10) alone may not fully prove that the regularized and unregularized losses share all minima, but that is an omitted proof, not a reduction of the result to its inputs. No circular step of the enumerated kinds is present.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The method introduces no new physical or model entity; it tunes alpha and uses existing prototypes. The main uncharged input is the correctness of the gradient-based motivation and the minimum-sharing assumption.

free parameters (2)
  • alpha (label-overlap reweighting exponent) = 0 or 1, selected per dataset by validation (Table 6)
    Controls f(y_i,y_j) in Eq. (17); the appendix states alpha should be treated as a tunable hyperparameter.
  • temperature tau = 0.1 in main text, 0.01 in Appendix A.2
    Temperature controls softmax sharpness and the activation frequency of the regularization; the paper gives conflicting values.
assumptions (3)
  • ad hoc to paper Positive pairs whose softmax score exceeds their normalized weight Lambda produce repulsive gradients that should be removed.
    Central motivation for the regularization; relies on gradient Eqs. (8)-(9) which appear incorrect.
  • ad hoc to paper The regularized loss and the original loss share the same minima.
    Claimed via Eq. (10), but Eq. (10) only characterizes stationary points of the original loss, not global minima of the regularized objective.
  • domain assumption Trainable label prototypes act as stable reference points for each label and can be treated as additional batch members.
    Adopted from prior work [20, 36]; not proved in this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-Label Contrastive Learning : A Comprehensive Study." pith.science (2026). https://pith.science/paper/2AQK7W4P

@misc{pith2026241200101,
  author       = {Pith},
  title        = {Pith review of: Multi-Label Contrastive Learning : A Comprehensive Study},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2AQK7W4P}},
  note         = {Machine review of arXiv:2412.00101}
}
read the original abstract

Multi-label classification, which involves assigning multiple labels to a single input, has emerged as a key area in both research and industry due to its wide-ranging applications. Designing effective loss functions is crucial for optimizing deep neural networks for this task, as they significantly influence model performance and efficiency. Traditional loss functions, which often maximize likelihood under the assumption of label independence, may struggle to capture complex label relationships. Recent research has turned to supervised contrastive learning, a method that aims to create a structured representation space by bringing similar instances closer together and pushing dissimilar ones apart. Although contrastive learning offers a promising approach, applying it to multi-label classification presents unique challenges, particularly in managing label interactions and data structure. In this paper, we conduct an in-depth study of contrastive learning loss for multi-label classification across diverse settings. These include datasets with both small and large numbers of labels, datasets with varying amounts of training data, and applications in both computer vision and natural language processing. Our empirical results indicate that the promising outcomes of contrastive learning are attributable not only to the consideration of label interactions but also to the robust optimization scheme of the contrastive loss. Furthermore, while the supervised contrastive loss function faces challenges with datasets containing a small number of labels and ranking-based metrics, it demonstrates excellent performance, particularly in terms of Macro-F1, on datasets with a large number of labels.

Figures

Figures reproduced from arXiv: 2412.00101 by the authors.

Figure 1
Figure 1. Visualization of gradient behavior in the multi-label contrastive loss: too close em [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Macro-F1 under limited data conditions as a function of training set size: Contrastive [PITH_FULL_IMAGE:figures/full_fig_p015_2.png] view at source ↗
Figure 3
Figure 3. Performance across training epochs with random initialization: Contrastive losses [PITH_FULL_IMAGE:figures/full_fig_p025_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 39 canonical work pages

  1. [1]

    Query2label: A simple transformer way to multi-label classification

    Shilong Liu, Lei Zhang, Xiao Yang, Hang Su, and Jun Zhu. Query2label: A simple transformer way to multi-label classification. arXiv preprint arXiv:2107.10834 , 2021

  2. [2]

    An exploration of encoder-decoder approaches to multi-label classification for legal and biomedical text

    Yova Kementchedjhieva and Ilias Chalkidis. An exploration of encoder-decoder approaches to multi-label classification for legal and biomedical text. In Anna Rogers, Jordan Boyd- Graber, and Naoaki Okazaki, editors, Findings of the Association for Computational Lin- guistics: ACL 2023 , pages 5828–5843, Toronto, Canada, July 2023. Association for Com- puta...

  3. [3]

    Large-scale multi-label text classification on EU legislation

    Ilias Chalkidis, Emmanouil Fergadiotis, Prodromos Malakasiotis, and Ion Androutsopou- los. Large-scale multi-label text classification on EU legislation. In Anna Korhonen, David Traum, and Llu ´ ıs M` arquez, editors,Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics , pages 6314–6322, Florence, Italy, July 2019. Assoc...

  4. [4]

    Emograph: Capturing emotion correlations using graph networks, 2020

    Peng Xu, Zihan Liu, Genta Indra Winata, Zhaojiang Lin, and Pascale Fung. Emograph: Capturing emotion correlations using graph networks, 2020

  5. [5]

    Label-representative graph convolutional network for multi-label text classification

    Huy-The Vu, Minh-Tien Nguyen, Van-Chien Nguyen, Minh-Hieu Pham, Van-Quyet Nguyen, and Van-Hau Nguyen. Label-representative graph convolutional network for multi-label text classification. Applied Intelligence, 2022

  6. [6]

    Multi-label image recogni- tion with graph convolutional networks

    Zhao-Min Chen, Xiu-Shen Wei, Peng Wang, and Yanwen Guo. Multi-label image recogni- tion with graph convolutional networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 5177–5186, 2019

  7. [7]

    SGM: sequence generation model for multi-label classification

    Pengcheng Yang, Xu Sun, Wei Li, Shuming Ma, Wei Wu, and Houfeng Wang. SGM: sequence generation model for multi-label classification. In Proceedings of the 27th Inter- national Conference on Computational Linguistics, COLING 2018, Santa Fe, New Mexico, USA, August 20-26, 2018 , 2018

  8. [8]

    Orderless recurrent models for multi-label classification

    Vacit Oguz Yazici, Abel Gonzalez-Garcia, Arnau Ramisa, Bartlomiej Twardowski, and Joost van de Weijer. Orderless recurrent models for multi-label classification. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 13440–13449, 2020

Show all 49 references
  1. [9]

    Enhancing label correlation feedback in multi-label text classification via multi-task learning

    Ximing Zhang, Qian-Wen Zhang, Zhao Yan, Ruifang Liu, and Yunbo Cao. Enhancing label correlation feedback in multi-label text classification via multi-task learning. arXiv preprint arXiv:2106.03103, 2021

  2. [10]

    An effective deploy- ment of contrastive learning in multi-label text classification

    Nankai Lin, Guanqiu Qin, Gang Wang, Dong Zhou, and Aimin Yang. An effective deploy- ment of contrastive learning in multi-label text classification. InFindings of the Association for Computational Linguistics: ACL 2023 , Toronto, Canada, July 2023. Association for Computationa...

  3. [11]

    Hierarchical multi-label classifica- tion networks

    Jonatas Wehrmann, Ricardo Cerri, and Rodrigo Barros. Hierarchical multi-label classifica- tion networks. In International conference on machine learning, pages 5075–5084. PMLR, 2018. 18

  4. [12]

    Hierarchy-aware label semantics matching network for hierarchical text classification

    Haibin Chen, Qianli Ma, Zhenxi Lin, and Jiangyue Yan. Hierarchy-aware label semantics matching network for hierarchical text classification. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference o...

  5. [13]

    Label-specific document representation for multi-label text classification

    Lin Xiao, Xin Huang, Boli Chen, and Liping Jing. Label-specific document representation for multi-label text classification. In Proceedings of the 2019 conference on empirical meth- ods in natural language processing and the 9th international joint conference on natural langua...

  6. [14]

    Ml- decoder: Scalable and versatile classification head

    Tal Ridnik, Gilad Sharir, Avi Ben-Cohen, Emanuel Ben-Baruch, and Asaf Noy. Ml- decoder: Scalable and versatile classification head. In Proceedings of the IEEE/CVF winter conference on applications of computer vision , pages 32–41, 2023

  7. [15]

    Focal loss for dense object detection

    Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Doll´ ar. Focal loss for dense object detection. In Proceedings of the IEEE international conference on computer vision, 2017

  8. [16]

    Asymmetric loss for multi-label classification

    Tal Ridnik, Emanuel Ben-Baruch, Nadav Zamir, Asaf Noy, Itamar Friedman, Matan Prot- ter, and Lihi Zelnik-Manor. Asymmetric loss for multi-label classification. In Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021

  9. [17]

    Zlpr: A novel loss for multi-label classification

    Jianlin Su, Mingren Zhu, Ahmed Murtadha, Shengfeng Pan, Bo Wen, and Yunfeng Liu. Zlpr: A novel loss for multi-label classification. arXiv preprint arXiv:2208.02955 , 2022

  10. [18]

    Exploring contrastive learning for long-tailed multi-label text classification

    Alexandre Audibert, Aur´ elien Gauffre, and Massih-Reza Amini. Exploring contrastive learning for long-tailed multi-label text classification. In Proceedings of the European Con- ference on Machine Learning and Knowledge Discovery in Databases (ECMLPKDD), pages 245–261, 2024

  11. [19]

    Multi-label supervised contrastive learning

    Pingyue Zhang and Mengyue Wu. Multi-label supervised contrastive learning. In Pro- ceedings of the AAAI Conference on Artificial Intelligence , volume 38, pages 16786–16793, 2024

  12. [20]

    Class prototypes based contrastive learning for classifying multi-label and fine-grained educational videos

    Rohit Gupta, Anirban Roy, Claire Christensen, Sujeong Kim, Sarah Gerard, Madeline Cincebeaux, Ajay Divakaran, Todd Grindal, and Mubarak Shah. Class prototypes based contrastive learning for classifying multi-label and fine-grained educational videos. In Proceedings of the IEEE...

  13. [21]

    A simple frame- work for contrastive learning of visual representations

    Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. A simple frame- work for contrastive learning of visual representations. In International conference on machine learning, pages 1597–1607. PMLR, 2020

  14. [22]

    Improving pairwise ranking for multi-label im- age classification

    Yuncheng Li, Yale Song, and Jiebo Luo. Improving pairwise ranking for multi-label im- age classification. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3617–3625, 2017

  15. [23]

    Momentum con- trast for unsupervised visual representation learning

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum con- trast for unsupervised visual representation learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 9729–9738, 2020

  16. [24]

    Supervised contrastive learning, 2021

    Prannay Khosla, Piotr Teterwak, Chen Wang, Aaron Sarna, Yonglong Tian, Phillip Isola, Aaron Maschinot, Ce Liu, and Dilip Krishnan. Supervised contrastive learning, 2021. 19

  17. [25]

    Dissecting su- pervised contrastive learning

    Florian Graf, Christoph Hofer, Marc Niethammer, and Roland Kwitt. Dissecting su- pervised contrastive learning. In International Conference on Machine Learning , pages 3821–3830. PMLR, 2021

  18. [26]

    Not all negatives are equal: Label-aware contrastive loss for fine-grained text classification

    Varsha Suresh and Desmond Ong. Not all negatives are equal: Label-aware contrastive loss for fine-grained text classification. In Marie-Francine Moens, Xuanjing Huang, Lucia Specia, and Scott Wen-tau Yih, editors, Proceedings of the 2021 Conference on Empiri- cal Methods in Na...

  19. [27]

    Label anchored contrastive learning for language understanding

    Zhenyu Zhang, Yuming Zhao, Meng Chen, and Xiaodong He. Label anchored contrastive learning for language understanding. In Marine Carpuat, Marie-Catherine de Marneffe, and Ivan Vladimir Meza Ruiz, editors, Proceedings of the 2022 Conference of the North American Chapter of the ...

  20. [28]

    Balanced contrastive learning for long-tailed visual recognition

    Jianggang Zhu, Zheng Wang, Jingjing Chen, Yi-Ping Phoebe Chen, and Yu-Gang Jiang. Balanced contrastive learning for long-tailed visual recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 6908–6917, 2022

  21. [29]

    Parametric contrastive learning

    Jiequan Cui, Zhisheng Zhong, Shu Liu, Bei Yu, and Jiaya Jia. Parametric contrastive learning. In Proceedings of the IEEE/CVF international conference on computer vision , pages 715–724, 2021

  22. [30]

    Contrastive learning-enhanced nearest neighbor mechanism for multi-label text classification

    Xi’ao Su, Ran Wang, and Xinyu Dai. Contrastive learning-enhanced nearest neighbor mechanism for multi-label text classification. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) , Dublin, Ireland, May 2022. Ass...

  23. [31]

    Disentangled variational autoencoder based multi-label classification with covariance-aware multivariate probit model

    Junwen Bai, Shufeng Kong, and Carla Gomes. Disentangled variational autoencoder based multi-label classification with covariance-aware multivariate probit model. In Christian Bessiere, editor, Proceedings of the Twenty-Ninth International Joint Conference on Ar- tificial Intel...

  24. [32]

    Use all the labels: A hierar- chical multi-label contrastive learning framework

    Shu Zhang, Ran Xu, Caiming Xiong, and Chetan Ramaiah. Use all the labels: A hierar- chical multi-label contrastive learning framework. In CVPR, 2022

  25. [33]

    Incorporating hierarchy into text encoder: a contrastive learning approach for hierarchical text classifi- cation

    Zihan Wang, Peiyi Wang, Lianzhe Huang, Xin Sun, and Houfeng Wang. Incorporating hierarchy into text encoder: a contrastive learning approach for hierarchical text classifi- cation. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volu...

  26. [34]

    Simon Chi Lok U, Jie He, V ´ ıctor Guti´ errez-Basulto, and Jeff Z. Pan. Instances and labels: Hierarchy-aware joint supervised contrastive learning for hierarchical multi-label text classification, 2023

  27. [35]

    Dao, Ethan Zhao, Dinh Phung, and Jianfei Cai

    Son D. Dao, Ethan Zhao, Dinh Phung, and Jianfei Cai. Multi-label image classification with contrastive learning, 2021. 20

  28. [36]

    A unified contrastive loss for self-training

    Aurelien Gauffre, Julien Horvat, and Massih-Reza Amini. A unified contrastive loss for self-training. In Machine Learning and Knowledge Discovery in Databases. Research Track and Demo Track, pages 3–18, Cham, 2024. Springer Nature Switzerland

  29. [37]

    Improved deep metric learning with multi-class n-pair loss objective

    Kihyuk Sohn. Improved deep metric learning with multi-class n-pair loss objective. In D. Lee, M. Sugiyama, U. Luxburg, I. Guyon, and R. Garnett, editors, Advances in Neural Information Processing Systems, volume 29. Curran Associates, Inc., 2016

  30. [38]

    The pascal visual object classes (voc) challenge

    Mark Everingham, Luc Van Gool, Christopher KI Williams, John Winn, and Andrew Zisserman. The pascal visual object classes (voc) challenge. International journal of computer vision, 88:303–338, 2010

  31. [39]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedin...

  32. [40]

    Nus-wide: a real-world web image database from national university of singapore

    Tat-Seng Chua, Jinhui Tang, Richang Hong, Haojie Li, Zhiping Luo, and Yantao Zheng. Nus-wide: a real-world web image database from national university of singapore. In Proceedings of the ACM international conference on image and video retrieval , pages 1–9, 2009

  33. [41]

    Rcv1-v2/lyrl2004: the lyrl2004 distribution of the rcv1-v2 text categoriza- tion test collection, 2004

    David D Lewis. Rcv1-v2/lyrl2004: the lyrl2004 distribution of the rcv1-v2 text categoriza- tion test collection, 2004

  34. [42]

    Hierarchical multi-label classification of text with capsule networks

    Rami Aly, Steffen Remus, and Chris Biemann. Hierarchical multi-label classification of text with capsule networks. In Fernando Alva-Manchego, Eunsol Choi, and Daniel Khashabi, editors, Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: Stu...

  35. [43]

    Roberta: A robustly optimized bert pretraining approach

    Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692 , 2019

  36. [44]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017

  37. [45]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

  38. [46]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248–255. Ieee, 2009

  39. [47]

    Understanding contrastive representation learning through alignment and uniformity on the hypersphere

    Tongzhou Wang and Phillip Isola. Understanding contrastive representation learning through alignment and uniformity on the hypersphere. In International conference on machine learning, pages 9929–9939. PMLR, 2020

  40. [48]

    Newsweeder: Learning to filter netnews

    Ken Lang. Newsweeder: Learning to filter netnews. In Armand Prieditis and Stuart Russell, editors, Machine Learning Proceedings 1995, pages 331–339. Morgan Kaufmann, San Francisco (CA), 1995. 21

  41. [49]

    Semeval-2018 task 1: Affect in tweets

    Saif Mohammad, Felipe Bravo-Marquez, Mohammad Salameh, and Svetlana Kiritchenko. Semeval-2018 task 1: Affect in tweets. In Proceedings of the 12th international workshop on semantic evaluation , pages 1–17, 2018. 22 A Appendix A.1 Implementation Details We train the standard l...

Pith tools

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