Pith. sign in

REVIEW 2 major objections 5 minor 53 references

Structural Entropy Guided Probabilistic Coding

T0 review · 2 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read SEPC claims that maximizing the structural entropy of a label-induced partition, applied to probabilistic embeddings, improves classification and regression on 12 natural-language tasks.

desk verdict Solid classification regularizer with a genuinely new structural-entropy objective, but the regression half is underspecified to the point of being unreproducible from the text. read the letter →

arxiv 2412.08841 v2 pith:FSFKJRO5 submitted 2024-12-12 cs.AI

classification cs.AI
keywords probabilisticcodingstructuralentropyembeddinginformationbottleneckregularizationnaturallanguageunderstandingregressionasclassificationlabelnoiserobustness
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 SEPC, a probabilistic coding method that adds a structural-entropy regularization term to the usual information-bottleneck objective. The claim is that maximizing the structural entropy of the partition induced by class labels separates the latent distributions of different classes and thereby improves generalization, robustness to label noise, and performance under scarce training data. For regression, the paper introduces a probabilistic encoding tree that turns continuous labels into soft probabilities over bin centers, allowing the same structural-entropy loss to be used. Experiments on 10 classification and 2 regression natural-language tasks report that SEPC outperforms deterministic and probabilistic baselines on both BERT and RoBERTa backbones. A sympathetic reader would take the paper to be establishing that inter-variable structural information, not just per-variable constraints, belongs in probabilistic embedding objectives.

What carries the argument

The load-bearing object is the structural-entropy regularization loss $L_{SE}$, built from the graph whose edge weights are sigmoid similarities between latent embeddings and whose assignment matrix $C$ encodes (soft) class membership. Structural entropy measures the information in a hierarchical partition of a graph by summing, over non-root nodes, cut-weight proportions times log volume ratios; here the three-tier encoding tree has classes as the intermediate layer. Maximizing the intermediate-layer structural entropy pulls different class distributions apart, and the loss is attached to the probabilistic coding loss as $L_{SEPC} = L_{PC} - \gamma L_{SE}$. For regression, the probabilistic encoding tree replaces the hard assignment matrix with the soft-label matrix $Y' = \operatorname{softmax}(-D)$, where $D$ holds distances to bin centers, so the same formula (Equation 14) applies to continuous labels.

What would settle it

Re-run SEPC's regression experiments after fixing a concrete decoding rule, for example taking the expected bin center under the predicted soft distribution, and check whether the reported Pearson and Spearman gains over SPC on STS-B and Claire survive; if the gains vanish or cannot be reproduced without an arbitrary decoding choice, the regression claim fails. Separately, replacing the label-based intermediate partition with random class assignments in classification should destroy most of the reported improvements if the structural-entropy mechanism is the cause.

Watch

Extended reading notes

Core claim

SEPC's central discovery is that structural entropy can serve as a regularizer in the opposite direction from prior usage: instead of minimizing the structural entropy of an encoding tree to compress task information, the model maximizes the structural entropy of a three-tier tree whose intermediate nodes are classes. This pushes the probabilistic embeddings of different classes apart in latent space while keeping the encoder-only probabilistic coding framework. The regularization loss is Equation (7) for hard labels and Equation (14) for soft labels, with both reducing to the same form once the assignment matrix is allowed to hold probabilities. The regression extension discretizes the label range into bin centers, forms soft labels by $Y' = \operatorname{softmax}(-D)$ on distances to those centers, and relaxes the encoding tree so each leaf can belong to every intermediate node with varying probability; the paper reports that this beats hard discretization and improves Pearson and Spearman correlations on STS-B and Claire.

Load-bearing premise

The regression recipe rests on the assumption that replacing a continuous label with a soft distribution over a few bin centers and later decoding the predicted bin distribution back to a real number loses no information the model needs, but the paper never specifies the output head or decoding rule.

Editorial extensions

If this is right

  • If SEPC's results hold, probabilistic embeddings can be regularized with inter-variable structure at no architectural cost, since the loss is computed from embeddings and labels already in the pipeline.
  • Regression benefits from classification-style structural regularization once labels are softened, so hard binning is not required to apply structural information theory to continuous outputs.
  • The largest reported gains on the Hate and Irony datasets indicate the regularizer helps most where train-test topic imbalance or subtle semantics make the task hard.
  • The 10%, 20%, and 30% label-noise experiments imply the class-level structural objective tolerates individual label flips better than per-sample objectives.
  • The method introduces one hyperparameter $\gamma$, and the reported sensitivity analysis says lower values are generally preferred and $\gamma=10$ hurts, so deployment is cheap.

Reading between the lines

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

  • One extension the paper leaves open is replacing the label-derived intermediate partition with a cluster-derived partition; if the gains persist, the regularizer is not tied to label information.
  • A reader may test the same soft-bin recipe on other continuous outputs, such as age estimation or relevance scoring, where the bin-center decoding rule would have to be chosen explicitly.
  • The robustness experiments use symmetric random label flips; class-conditional or structured noise would better test whether structural entropy protects against realistic annotation errors.
  • The paper never specifies how predicted bin probabilities become a scalar, so any reimplementation must fix that rule; the reported regression gains should survive whichever reasonable rule is chosen.
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

2 major / 5 minor

Summary. The paper proposes SEPC, a probabilistic coding model with a structural-entropy-based regularization loss. For classification, the model constructs a graph over latent embeddings, builds a three-tier encoding tree from class labels, and maximizes the structural entropy of the class partition (Eqs. 5-8). For regression, the paper proposes a 'probabilistic encoding tree': continuous labels are binned and softened into a distribution over bin centers (Eqs. 9-10), and the same structural-entropy loss is applied with soft assignments (Eq. 14). Experiments on 10 classification and 2 regression NLU tasks with BERT and RoBERTa backbones report consistent improvements over deterministic and probabilistic baselines, with ablations (w/o SE, w/o soft), label-noise robustness, limited-data generalization, and hyperparameter sensitivity studies.

Significance. The central idea---using the maximization of structural entropy as a regularizer for probabilistic embeddings---is novel and the empirical results are encouraging. The paper reports means and standard deviations over five seeds, includes ablations that isolate the structural-entropy term in classification, provides a robustness analysis, and links to public code, which are all strengths. If the regression pipeline is fully specified, the second contribution (probabilistic encoding tree for soft labels) would also be a useful practical technique. However, as written, the regression method is not reproducible from the text, and the claimed regression gains in Table 2 rest on an underspecified component.

major comments (2)
  1. [Probabilistic Encoding Tree for Regression Tasks, Eqs. (9)-(14)] The regression pipeline is underspecified. The paper defines soft labels Y' = softmax(-D) and the regularizer L_SE in Eq. (14), but it never states (i) the model's output head for regression (e.g., a linear layer with r logits), (ii) the task loss used to train against Y' (e.g., cross-entropy) or whether a scalar MSE head is retained with L_SE as an auxiliary regularizer, or (iii) the decoding rule from the predicted bin distribution to a continuous value (e.g., expectation over bin centers, argmax bin center, or a learned mapping). The bin centers P in Eq. (9) are also not specified beyond 'bin the entire regression label value space into r classes.' Consequently, Table 2 cannot be reimplemented from the text, and the second stated contribution ('effective method to utilize structural entropy for regression tasks') is not fully defined. Please specify these components explicitly, or state that the code is the only complete specification and summarize the details in the paper.
  2. [Probabilistic Encoding Tree for Regression Tasks, Eqs. (11)-(14)] The relationship between the verbal definition of the soft cut weight g'_alpha_j and the matrix expression in Eq. (14) is not derived. The text says the weight of cut edges 'should be multiplied by the probability of one vertex belonging to T_alpha_j and the other belonging to T^complement_alpha_j,' but Eq. (14) computes [(1-C)^T A C]_jj. For symmetric A this equals sum_{i,k} C_{ij} A_{ik}(1-C_{kj}), which is a particular soft-cut convention; please show that this matches the stated definition and discuss how the relaxation affects the encoding-tree properties from the Preliminaries (in particular, the single-parent assignment of leaf nodes). Without this derivation, the probabilistic encoding tree is presented as an ad hoc formula rather than a principled extension of structural entropy.
minor comments (5)
  1. [Hyperparameter Sensitivity Analysis (Figure 4) and Generalization Analysis (Figure 5)] Figures 4 and 5 do not include error bars or confidence intervals, although the tables report standard deviations over five seeds; adding them would help assess the stability of the sensitivity and generalization results.
  2. [Eq. (10)] The softmax in Eq. (10) implicitly uses unit temperature; the temperature controls the softness of the soft labels and is not mentioned as a tuned or fixed choice. Please state the temperature or discuss its effect.
  3. [Eq. (9) and Parameter Settings] The choice of bin centers P is not described. For STS-B (labels 0-5, r=5) and Claire (labels 1-5, r=4), please specify whether equal-width bins are used and how the centers are computed.
  4. [Classification ablation (Table 1)] The text says SEPC w/o SE 'is the same as SPC w/o S model (Hu et al. 2024)' but then notes different hyperparameters and environments; please clarify the exact relationship so that the ablation cleanly isolates the contribution of L_SE rather than differences in training setup.
  5. [Related Work] There is a minor typo in the Related Work section: 'the conditional entropy bottleneck top improved robustness to adversarial examples' should likely read 'the conditional entropy bottleneck to improve robustness to adversarial examples.'

Circularity Check

0 steps flagged · score 2.0 of 10

No circular reduction found: the structural-entropy terms are supervised regularizers, and the cited self-works are contextual; the regression decoding gap is a reproducibility issue rather than a circularity.

full rationale

The paper's formal chain is self-contained as an optimization method: Eq. 7 is the structural entropy of a three-tier label tree instantiated from Li and Pan's definition (Eqs. 3-4), and Eq. 14 is the same formula with the soft assignment C=Y'=softmax(-D). Both are supervised regularizers added to L_PC (Eq. 8), not predictions derived from their own inputs. The empirical claims rest on comparisons with independent baselines (VIB, MINE-IB, MEIB, SPC) and on ablations; beta and gamma are searched hyperparameters, not fitted quantities renamed as predictions. The structural-entropy literature cited includes several papers by the current group (e.g., Zou et al. 2023; Zeng, Peng, and Li 2023; Sun et al. 2024; Peng et al. 2024), but these are contextual related-work citations, not load-bearing support for the method. The one substantive defect is that the regression decoding rule is absent: after defining Y' and L_SE (Eqs. 9-14), the paper never specifies how predicted bin probabilities are converted to a continuous score, so Table 2 is not reimplementable from the text. That is a reproducibility and correctness gap, not a circular reduction, so the circularity score is low.

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

The method relies on the standard IB/Gaussian assumptions plus a few ad hoc modeling choices: the sigmoid graph construction, the use of labels as the partition, and the fixed-temperature soft label for regression. The free parameters are hyperparameters and binning choices, all selected by validation or dataset-specific heuristics.

free parameters (4)
  • beta = searched from {1e-2,1e-1,1,10}
    Trade-off parameter in the IB objective L_PC (Eq. 2), selected per dataset by validation search.
  • gamma = searched from {1e-2,1e-1,1,10}
    Weight of the structural entropy loss LSE (Eq. 8), selected per dataset by validation search.
  • r (number of regression bins) = 5 for STS-B, 4 for Claire
    Chosen based on the label range (0-5 and 1-5); affects the granularity of the soft labels.
  • bin centers P = not specified, implied by binning the label space
    The paper says labels are binned into r classes but does not state whether bins are equal-width or data-driven; this choice affects D and hence Y'.
assumptions (5)
  • domain assumption The probabilistic coding objective is derived from the Markov chain Y -> X -> Z and the IB Lagrangian (Eq. 1).
    The paper inherits this from Tishby et al. and Alemi et al.; it is a standard assumption in IB-based representation learning, not justified anew.
  • domain assumption The latent variable z follows a Gaussian distribution p(z|x) = N(z; mu, Sigma) with the reparameterization trick.
    Assumed in 'Probabilistic Coding' section; standard for VIB-style models.
  • domain assumption Class labels constitute an optimal partition of the data for constructing the encoding tree.
    The paper states 'We treat the labels as the optimal partition for the data' in the Proposed Method; if labels do not correspond to separable clusters, the structural entropy regularizer may push embeddings toward incorrect separation.
  • ad hoc to paper The graph adjacency A = sigmoid(HZ HZ^T) (Eq. 5) faithfully represents the structural relationships between latent variables.
    The sigmoid dot-product similarity is a design choice without theoretical justification; the behavior of LSE depends on this particular nonlinearity and on batch-level computation.
  • ad hoc to paper For regression, the soft label distribution Y' = softmax(-D) (Eq. 10) with unit temperature is a suitable target for classification-style training.
    The temperature is fixed at 1, so the softness depends on the arbitrary scale of the label values; no tuning or theoretical basis is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Structural Entropy Guided Probabilistic Coding." pith.science (2026). https://pith.science/paper/FSFKJRO5

@misc{pith2026241208841,
  author       = {Pith},
  title        = {Pith review of: Structural Entropy Guided Probabilistic Coding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FSFKJRO5}},
  note         = {Machine review of arXiv:2412.08841}
}
read the original abstract

Probabilistic embeddings have several advantages over deterministic embeddings as they map each data point to a distribution, which better describes the uncertainty and complexity of data. Many works focus on adjusting the distribution constraint under the Information Bottleneck (IB) principle to enhance representation learning. However, these proposed regularization terms only consider the constraint of each latent variable, omitting the structural information between latent variables. In this paper, we propose a novel structural entropy-guided probabilistic coding model, named SEPC. Specifically, we incorporate the relationship between latent variables into the optimization by proposing a structural entropy regularization loss. Besides, as traditional structural information theory is not well-suited for regression tasks, we propose a probabilistic encoding tree, transferring regression tasks to classification tasks while diminishing the influence of the transformation. Experimental results across 12 natural language understanding tasks, including both classification and regression tasks, demonstrate the superior performance of SEPC compared to other state-of-the-art models in terms of effectiveness, generalization capability, and robustness to label noise. The codes and datasets are available at https://github.com/SELGroup/SEPC.

Figures

Figures reproduced from arXiv: 2412.08841 by the authors.

Figure 1
Figure 1. Two common architectures of probabilistic coding. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overall model of SEPC. (1) Each tree node α corresponds to a subset of data points Tα ⊆ X. Especially, for the root node λ of T , we define the points set it associated with as Tλ = X. For the leaf node α at the last depth, Tα is a singleton containing a single data point x ∈ X. If the leaf node α is not at the last depth, Tα is ∅. (2) For each non-leaf tree node α, its i-th immediate child is α <i>, and its par… view at source ↗
Figure 4
Figure 4. The impact of the weight parameter γ on the regularization loss LSE. tic encoding tree. Instead, we assign each sample to the class with the closest distance to the class bin center and use the normal encoding tree to calculate LSE. As shown in [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Results of different models with different ratios of the training set. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Visualization of embeddings. The circle represents [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 43 canonical work pages

  1. [1]

    A.; Fischer, I.; Dillon, J

    Alemi, A. A.; Fischer, I.; Dillon, J. V.; and Murphy, K. 2017. Deep Variational Information Bottleneck . In International Conference on Learning Representations ( ICLR )

  2. [2]

    An, S.; Jammalamadaka, N.; and Chong, E. 2023. Maximum Entropy Information Bottleneck for Uncertainty -aware Stochastic Embedding . In Computer Vision and Pattern Recognition ( CVPR ) , 3809--3818

  3. [3]

    E.; Ballesteros, M.; Basile, V.; Patti, V.; and Saggion, H

    Barbieri, F.; Camacho-Collados, J.; Ronzano, F.; Anke, L. E.; Ballesteros, M.; Basile, V.; Patti, V.; and Saggion, H. 2018. Semeval 2018 task 2: Multilingual emoji prediction. In Proceedings of the 12th international workshop on semantic evaluation, 24--33

  4. [4]

    Basile, V.; Bosco, C.; Fersini, E.; Nozza, D.; Patti, V.; Pardo, F. M. R.; Rosso, P.; and Sanguinetti, M. 2019. Semeval-2019 task 5: Multilingual detection of hate speech against immigrants and women in twitter. In Proceedings of the 13th international workshop on semantic evaluation, 54--63

  5. [5]

    I.; Baratin, A.; Rajeswar, S.; Ozair, S.; Bengio, Y.; Hjelm, R

    Belghazi, M. I.; Baratin, A.; Rajeswar, S.; Ozair, S.; Bengio, Y.; Hjelm, R. D.; and Courville, A. C. 2018. Mutual Information Neural Estimation . In International Conference on Machine Learning , 530--539

  6. [6]

    Cer, D.; Diab, M.; Agirre, E.; Lopez-Gazpio, I.; and Specia, L. 2017. S em E val-2017 Task 1: Semantic Textual Similarity Multilingual and Crosslingual Focused Evaluation. In Proceedings of the 11th International Workshop on Semantic Evaluation ( S em E val-2017) , 1--14

  7. [7]

    Chalk, M.; Marre, O.; and Tkacik, G. 2016. Relevant sparse codes with variational information bottleneck. In Conference on Neural Information Processing Systems ( NeurIPS ) , 1957--1965

  8. [8]

    Cortes, C.; and Vapnik, V. 1995. Support-vector networks. Machine learning, 20: 273--297

Show all 53 references
  1. [9]

    Demszky, D.; Movshovitz-Attias, D.; Ko, J.; Cowen, A.; Nemade, G.; and Ravi, S. 2020. G o E motions: A Dataset of Fine-Grained Emotions. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 4040--4054

  2. [10]

    Devlin, J.; Chang, M.-W.; Lee, K.; and Toutanova, K. 2019. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding . In North American Chapter of the Association for Computational Linguistics ( NAACL ) , 4171--4186

  3. [11]

    Dong, W.; Yan, D.; and Wang, P. 2024. Self-Supervised Node Representation Learning via Node-to-Neighbourhood Alignment. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46: 4218--4233

  4. [12]

    Duan, L.; Chen, X.; Liu, W.; Liu, D.; Yue, K.; and Li, A. 2024. Structural Entropy Based Graph Structure Learning for Node Classification . In AAAI Conference on Artificial Intelligence ( AAAI ) , 8372--8379

  5. [13]

    Fischer, I. 2020. The Conditional Entropy Bottleneck . Entropy, 22(9): 999

  6. [14]

    Gunel, B.; Du, J.; Conneau, A.; and Stoyanov, V. 2021. Supervised Contrastive Learning for Pre-trained Language Model Fine-tuning. In 9th International Conference on Learning Representations

  7. [15]

    Hochreiter, S.; and Schmidhuber, J. 1997. Long short-term memory. Neural computation, 9(8): 1735--1780

  8. [16]

    Hu, D.; Hou, X.; Du, X.; Zhou, M.; Jiang, L.; Mo, Y.; and Shi, X. 2022. VarMAE: Pre-training of Variational Masked Autoencoder for Domain -adaptive Language Understanding . In Conference on Empirical Methods in Natural Language Processing ( EMNLP ) , 6276--6286

  9. [17]

    Hu, D.; Wei, L.; Liu, Y.; Zhou, W.; and Hu, S. 2024. Structured Probabilistic Coding . In AAAI Conference on Artificial Intelligence ( AAAI ) , 12491--12501

  10. [18]

    Joulin, A.; Grave, E.; Bojanowski, P.; and Mikolov, T. 2017. Bag of Tricks for Efficient Text Classification. In Proceedings of the 15th Conference of the E uropean Chapter of the Association for Computational Linguistics: Volume 2, Short Papers , 427--431

  11. [19]

    Kim, J.; Kim, M.; Woo, D.; and Kim, G. 2021. Drop- Bottleneck : Learning Discrete Compressed Representation for Noise - Robust Exploration . In International Conference on Learning Representations ( ICLR )

  12. [20]

    P.; and Welling, M

    Kingma, D. P.; and Welling, M. 2013. Auto- Encoding Variational Bayes . In International Conference on Learning Representations . Explorandum Ltd

  13. [21]

    Li, A.; and Pan, Y. 2016. Structural Information and Dynamical Complexity of Networks . IEEE Transactions on Information Theory, 62(6): 3290--3339

  14. [22]

    Liu, Y.; Ott, M.; Goyal, N.; Du, J.; Joshi, M.; Chen, D.; Levy, O.; Lewis, M.; Zettlemoyer, L.; and Stoyanov, V. 2020. RoBERTa: A Robustly Optimized BERT Pretraining Approach . ArXiv, abs/1907.11692

  15. [23]

    Ma, J.; Wang, C.; Liu, Y.; Lin, L.; and Li, G. 2023. Enhanced Soft Label for Semi-Supervised Semantic Segmentation. 2023 IEEE/CVF International Conference on Computer Vision (ICCV), 1185--1195

  16. [24]

    K.; Belinkov, Y.; and Henderson, J

    Mahabadi, R. K.; Belinkov, Y.; and Henderson, J. 2021. Variational Information Bottleneck for Effective Low - Resource Fine - Tuning . In International Conference on Learning Representations ( ICLR )

  17. [25]

    M.; and Goodfellow, I

    Miyato, T.; Dai, A. M.; and Goodfellow, I. J. 2017. Adversarial Training Methods for Semi-Supervised Text Classification. In 5th International Conference on Learning Representations

  18. [26]

    Mohammad, S.; Bravo-Marquez, F.; Salameh, M.; and Kiritchenko, S. 2018. Semeval-2018 task 1: Affect in tweets. In Proceedings of the 12th international workshop on semantic evaluation, 1--17

  19. [27]

    Mohammad, S.; Kiritchenko, S.; Sobhani, P.; Zhu, X.; and Cherry, C. 2016. Semeval-2016 task 6: Detecting stance in tweets. In Proceedings of the 10th international workshop on semantic evaluation (SemEval-2016), 31--41

  20. [28]

    Muthukumar, V.; Narang, A.; Subramanian, V.; Belkin, M.; Hsu, D.; and Sahai, A. 2021. Classification vs regression in overparameterized regimes: Does the loss function matter? Journal of Machine Learning Research, 22(222): 1--69

  21. [29]

    J.; Murphy, K

    Oh, S. J.; Murphy, K. P.; Pan, J.; Roth, J.; Schroff, F.; and Gallagher, A. C. 2019. Modeling Uncertainty with Hedged Instance Embeddings . In International Conference on Learning Representations ( ICLR )

  22. [30]

    Peng, H.; Zhang, J.; Huang, X.; Hao, Z.; Li, A.; Yu, Z.; and Yu, P. S. 2024. Unsupervised Social Bot Detection via Structural Information Theory. TOIS, 42(6)

  23. [31]

    Pereyra, G.; Tucker, G.; Chorowski, J.; Kaiser, L.; and Hinton, G. E. 2017. Regularizing Neural Networks by Penalizing Confident Output Distributions. In 5th International Conference on Learning Representations

  24. [32]

    L.; Lin, Y.; Dijkstra, J.; and van Gemert, J

    Pintea, S. L.; Lin, Y.; Dijkstra, J.; and van Gemert, J. C. 2023. A step towards understanding why classification helps regression. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 19972--19981

  25. [33]

    Poria, S.; Hazarika, D.; Majumder, N.; Naik, G.; Cambria, E.; and Mihalcea, R. 2019. MELD : A Multimodal Multi-Party Dataset for Emotion Recognition in Conversations. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, 527--536

  26. [34]

    Rosenthal, S.; Farra, N.; and Nakov, P. 2017. S em E val-2017 Task 4: Sentiment Analysis in T witter. In Proceedings of the 11th International Workshop on Semantic Evaluation ( S em E val-2017) , 502--518

  27. [35]

    Roth, M.; Anthonio, T.; and Sauer, A. 2022. S em E val-2022 Task 7: Identifying Plausible Clarifications of Implicit and Underspecified Phrases in Instructional Texts. In Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022), 1039--1049

  28. [36]

    R.; and Wallbott, H

    Scherer, K. R.; and Wallbott, H. G. 1994. Evidence for universality and cultural variation of differential emotion response patterning. Journal of personality and social psychology, 66(2): 310

  29. [37]

    Shi, Y.; and Jain, A. K. 2019. Probabilistic Face Embeddings . In IEEE International Conference on Computer Vision ( ICCV ) , 6901--6910

  30. [38]

    Stewart, L.; Bach, F.; Berthet, Q.; and Vert, J.-P. 2023. Regression as classification: Influence of task formulation on neural network features. In International Conference on Artificial Intelligence and Statistics, 11563--11582. PMLR

  31. [39]

    Sun, L.; Huang, Z.; Peng, H.; Wang, Y.; Liu, C.; and Yu, P. S. 2024. LSEnet: Lorentz Structural Entropy Neural Network for Deep Graph Clustering . In Forty-first International Conference on Machine Learning

  32. [40]

    Sun, Q.; Li, J.; Peng, H.; Wu, J.; Fu, X.; Ji, C.; and Yu, P. S. 2022. Graph Structure Learning with Variational Information Bottleneck . In AAAI Conference on Artificial Intelligence ( AAAI ) , 4165--4174

  33. [41]

    C.; and Bialek, W

    Tishby, N.; Pereira, F. C.; and Bialek, W. 2000. The information bottleneck method. ArXiv, physics/0004057

  34. [42]

    Tishby, N.; and Zaslavsky, N. 2015. Deep learning and the information bottleneck principle. In Information Theory Workshop , 1--5

  35. [43]

    Van Hee, C.; Lefever, E.; and Hoste, V. 2018. Semeval-2018 task 3: Irony detection in english tweets. In Proceedings of the 12th international workshop on semantic evaluation, 39--50

  36. [44]

    Vilnis, L.; and McCallum, A. 2015. Word Representations via Gaussian Embedding . In International Conference on Learning Representations ( ICLR )

  37. [45]

    Wang, Y.; Wang, Y.; Zhang, Z.; Yang, S.; Zhao, K.; and Liu, J. 2023. USER: Unsupervised Structural Entropy - Based Robust Graph Neural Network . In AAAI Conference on Artificial Intelligence ( AAAI ) , 10235--10243. Association for the Advancement of Artificial Intelligence (AAAI)

  38. [46]

    Wu, J.; Chen, X.; Shi, B.; Li, S.; and Xu, K. 2023. SEGA: Structural Entropy Guided Anchor View for Graph Contrastive Learning . In International Conference on Machine Learning ( ICML ) , volume abs/2305.04501, 37293--37312

  39. [47]

    Xu, Z.; Wu, D.; Yu, C.; Chu, X.; Sang, N.; and Gao, C. 2024. SCTNet: Single- Branch CNN with Transformer Semantic Information for Real - Time Segmentation . In AAAI Conference on Artificial Intelligence ( AAAI ) , 6378--6386

  40. [48]

    Zampieri, M.; Malmasi, S.; Nakov, P.; Rosenthal, S.; Farra, N.; and Kumar, R. 2019. S em E val-2019 Task 6: Identifying and Categorizing Offensive Language in Social Media ( O ffens E val). In Proceedings of the 13th International Workshop on Semantic Evaluation, 75--86

  41. [49]

    Zeng, X.; Peng, H.; and Li, A. 2023. Effective and Stable Role - Based Multi - Agent Collaboration by Structural Information Principles . In AAAI Conference on Artificial Intelligence ( AAAI ) , 11772--11780

  42. [50]

    Zeng, X.; Peng, H.; and Li, A. 2024. Adversarial socialbots modeling based on structural information principles. In Proceedings of the AAAI, volume 38, 392--400

  43. [51]

    Zou, D.; Peng, H.; Huang, X.; Yang, R.; Li, J.; Wu, J.; Liu, C.; and Yu, P. S. 2023. SE- GSL : A General and Effective Graph Structure Learning Framework through Structural Entropy Optimization . In Proceedings of the ACM Web Conference 2023

  44. [52]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  45. [53]

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