Pith. sign in

REVIEW 3 major objections 4 minor 63 references

Hyperbolic Residual Quantization: Discrete Representations for Data with Latent Hierarchies

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper claims that moving residual quantization into hyperbolic space produces discrete multitoken representations that preserve latent hierarchies better than Euclidean residual quantization, with up to 20% gains on a noun hypernym…

desk verdict HRQ is a plausible incremental extension of RQ-VAE to hyperbolic space, but the paper as written hides the curvature and uses a Euclidean codebook loss, so the main comparison isn't reproducible. read the letter →

arxiv 2505.12404 v1 pith:XH4L466F submitted 2025-05-18 cs.LG cs.AI

Piotr Piękos , Subhradeep Kayal , Alexandros Karatzoglou This is my paper · ORCID
classification cs.LGcs.AI
keywords residualquantizationhyperbolicspacePoincaréballdiscreterepresentationlearningmultitokenhierarchicaldataRQ-VAElatenthierarchy
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 proposes Hyperbolic Residual Quantization (HRQ), which moves residual quantization into the Poincaré ball, a model of hyperbolic space. In HRQ, the encoder, the residual difference, the reconstruction sum, and the nearest-codebook search all use hyperbolic operations instead of Euclidean ones. The intended effect is to give discrete multitoken representations an inductive bias toward hierarchical branching, so that tokens naturally group by shared ancestors. On a noun hypernym prediction task, HRQ tokens improve Recall@10 by up to 20% over Euclidean RQ tokens, and on recommender benchmarks with no hierarchy supervision HRQ-VAE tokens are also better than RQ-VAE tokens. The takeaway is that the geometry of the quantization space, not just the training loss, controls how well discrete codes capture latent hierarchy.

What carries the argument

The central object is the Poincaré ball model of hyperbolic space with curvature $c$, a ball-shaped manifold whose volume grows exponentially with distance from the origin. The mechanism is the replacement of Euclidean subtraction, addition, and distance in residual quantization by Möbius subtraction $x\ominus_c y = x\oplus_c(-y)$, Möbius addition $x\oplus_c y$, and the hyperbolic distance $d_{\mathcal{P}_c}(u,v)=\operatorname{arcosh}(1+2c\|u-v\|^2/((1-c\|u\|^2)(1-c\|v\|^2)))$, which keeps every quantization step inside the manifold.

What would settle it

Train HRQ and RQ on the noun hypernym task while sweeping the Poincaré curvature $c$ across a wide range (e.g., $0.1$ to $10$) and report Recall@10 for each $c$; if no choice of $c$ beats Euclidean RQ, or if the gain appears only in a narrow window, the reported advantage depends on a favorable fixed constant rather than on hyperbolic geometry itself.

Watch

Extended reading notes

Core claim

HRQ-VAE replaces each component of residual quantization with a hyperbolic analogue: encoder and decoder are hyperbolic neural networks in the Poincaré ball, residuals are computed by Möbius subtraction, the quantized reconstruction is built by Möbius addition, and the nearest codebook vector is selected by hyperbolic distance. The training objective remains the standard autoencoder plus codebook loss of RQ-VAE, but optimization runs with a Riemannian optimizer. The paper's claim is that this geometric change gives the model an inductive bias toward hierarchical branching, so the resulting multitokens are more semantically organized and more useful to a downstream transformer, even when no hierarchy label is available during training.

Load-bearing premise

The load-bearing premise is that hyperbolic geometry is the right inductive bias for the latent hierarchies in these datasets, and that bias is controlled by the curvature $c$, which the paper never tunes or reports, so if the fixed $c$ mismatches the branching structure the claimed improvements could shrink or reverse.

Editorial extensions

If this is right

  • HRQ multitokens improve Recall@10 for hypernym generation over RQ multitokens by up to about 20%, across the tested codebook sizes, token lengths, and embedding dimensions.
  • On hierarchy discovery, HRQ-VAE multitokens beat RQ-VAE multitokens on next-item prediction metrics without any hierarchy supervision, on both product and movie datasets.
  • The downstream transformer architecture and training are identical for both token types in the hierarchy modeling experiments, so the performance gap is attributed to the tokens themselves.
  • HRQ produces a latent space with lower norm variability than RQ, which the paper interprets as a more tree-like spanning structure for quantization.
  • Because the training objective is unchanged, HRQ can be adopted wherever RQ-VAE is used, provided the data have latent hierarchies.

Reading between the lines

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

  • An untested extension suggested by the paper's structure is per-dataset curvature selection: if the Poincaré curvature $c$ is tuned, gains on deeper hierarchies could exceed the reported 20%.
  • The same hyperbolic residual mechanism could plausibly transfer to image and audio RQ-VAEs, which the paper explicitly leaves out of scope.
  • If HRQ's radial token structure is interpretable, shared prefix tokens should align with shared ancestors; counting token sharing across known hyponym pairs would test that directly.
  • A non-hierarchical control dataset would clarify whether the geometry helps specifically through hierarchies or through a generic regularization effect.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes Hyperbolic Residual Quantization (HRQ), a modification of RQ-VAE in which the latent space is the Poincaré ball, codebook selection uses hyperbolic distance, residuals are computed with Möbius subtraction, and reconstruction uses Möbius addition. The authors evaluate HRQ on WordNet hypernym modeling and on recommendation tasks (Amazon Beauty/Toys/Sports, MovieLens), reporting consistent gains over Euclidean RQ multitokens, with up to 20% relative improvement in hierarchy modeling. Appendix D offers a structural explanation based on reduced norm variation in hyperbolic space.

Significance. If the empirical claims hold, the paper fills a clear gap: discrete residual quantization in hyperbolic space has been mostly unexplored, and the multitoken representations it produces could benefit generative retrieval and interpretability. The method is simple, and the experimental design of keeping the downstream model fixed while changing only the tokenization is appropriate for isolating the effect of hyperbolic geometry. The paper is not circular: the reported gains are measured against an external Euclidean RQ baseline. However, the contribution's value is contingent on resolving the specification gaps below; the current manuscript does not support a definitive causal attribution to hyperbolic geometry.

major comments (3)
  1. [§3.1 and Algorithm 1] Algorithm 1's codebook loss l_cmt is written with Euclidean squared norms, ||sg[r_i^C]−e_i^C||^2 + α||r_i^C−sg[e_i^C]||^2, while §3.1 states that codebook selection uses the hyperbolic distance d_Pc of Eq. (1) and residual computation uses Möbius subtraction. If the implemented loss is Euclidean, the codebook is not trained with the hyperbolic metric and the 'full hyperbolic' claim in §3.1 is not supported; if the implemented loss is hyperbolic, Algorithm 1 is incorrect. This mismatch is load-bearing because the codebook geometry is exactly what the paper credits for the gains. Please state explicitly which loss is used, and if it is Euclidean, provide the hyperbolic version or an argument that the Euclidean norm on Poincaré coordinates is a sufficient training signal.
  2. [§2.1, §3, Appendix C] The Poincaré curvature c is never stated for any experiment. Eq. (1), the exponential/logarithmic maps, the Möbius operations, and Algorithm 1 all depend on c; as c→0 the hyperbolic operations converge to Euclidean ones, so without c the reader cannot tell whether HRQ's reported gains reflect hyperbolic geometry or an unreported, possibly dataset-specific choice of c. Appendix C gives learning rates, epochs, and architecture but no c value and no sensitivity analysis. In addition, Eq. (1) omits the standard 1/sqrt(c) prefactor of the Poincaré distance; while this does not change nearest-neighbor selection, it changes the scale of distances used in the §4.1 contrastive loss and should be corrected or explicitly justified. Report c for every experiment and add at least a small sweep (e.g., c in {0.01, 0.1, 1}) on the WordNet task to show that the claimed advantage is not an artifact of a single curvature.
  3. [§4.1, Tables 1 and 2] The statistical evidence is incompletely reported. Table 1 has no error bars, number of seeds, or significance tests, although 8–22% relative gains are claimed; Table 2 reports only mean±std over 8 runs, and several HRQ improvements (e.g., AR SaO NDCG@10, −1.4%; MovieLens NDCG@10, +2.7%) are within one standard deviation of the RQ baseline. Please report standard deviations or confidence intervals for the hierarchy-modeling task and run paired significance tests across repeated tokenization and downstream training to support the 'consistently outperform' claim.
minor comments (4)
  1. [Algorithm 1] The residual update ri+1_C ← ri_C ⊖_c e^{i−1}_c should use e^i_C rather than e^{i−1}_c, and Algorithm 1 should list c as an explicit input since all operations depend on it.
  2. [§1] The sentence 'In section 5. Finally, in Section 6 we summarize our findings' is incomplete; Section 5 is not introduced.
  3. [Throughout] There are several typos: 'Specficially' (§4.1), 'joinlty' (§4.1), 'taoxnomy' (§4.1), 'equiped' (Appendix C), 'parantheses' (Table 2 caption), and 'concanted' (Appendix C).
  4. [Appendix C.2] The text says the (H)RQ-VAE learning rate is chosen using the validation split of the downstream task; please clarify whether this selection happens before or after the downstream model is trained, since using downstream validation to select the tokenizer may affect the cleanliness of the comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: HRQ's improvements are empirical results measured against an external Euclidean RQ baseline, not derived from or equivalent to the paper's own inputs.

full rationale

The paper's central claim is that hyperbolic residual quantization yields more useful discrete hierarchical representations than Euclidean RQ. This is supported by measured downstream-task performance (Table 1 for WordNet hypernym modeling, Table 2 for hierarchy discovery on Amazon and MovieLens), where HRQ and RQ are compared on the same held-out tasks. The hyperbolic machinery (Poincaré ball, Möbius operations, hyperbolic distance, exponential/logarithmic maps) is standard and cited to independent external work (Gromov, Ganea et al., Nickel and Kiela), not to a self-citation chain by the present authors. No tested constant is fitted to the reported metric and then renamed as a prediction: the gains are observed outcomes of separately trained sequence-to-sequence and recommender models, and the multitoken creation is trained on an 85% split with the 15% test relations held out. Appendix D's norm analysis is post-hoc interpretation rather than a fitted input to the main result. The unreported curvature c and the Euclidean form of the codebook loss in Algorithm 1 are reproducibility or correctness concerns, not circular reductions of the claimed result to its own assumptions. Therefore the derivation chain is self-contained with respect to circularity and the appropriate score is 0.

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

The central claim is empirical, so the ledger mostly records experimental hyperparameters and implicit modeling assumptions. The load-bearing assumptions are the fixed curvature and the unstated replacement of hyperbolic distance by Euclidean norm in the codebook loss.

free parameters (4)
  • Poincaré curvature c = not reported (implicitly 1)
    Controls the exponential growth of space and hence the strength of the hierarchical inductive bias; never tuned or reported in Sections 2, 3, or Appendix C.
  • codebook size s = 64, 128, 256 (hierarchy modeling); 256 (hierarchy discovery)
    Swept in Table 1; fixed at 256 for the recommender experiments.
  • token length k = 3 and 4 (hierarchy modeling)
    Swept in Table 1; not reported for the recommender multitoken length.
  • RQ loss weight alpha = not reported
    Balances codebook-to-residual versus residual-to-codebook updates; no value or tuning procedure is given in Appendix C.
assumptions (5)
  • standard math Möbius addition and subtraction, exponential/log maps, and hyperbolic distance on the Poincaré ball are valid analogues of vector space operations for training neural networks
    Assumed from Ganea et al. [16] and Ungar [52] in Section 2.1.
  • standard math Hyperbolic space can embed tree-like data with low distortion, so it provides a useful inductive bias for latent hierarchies
    Invoked in Section 2.1 citing Gromov [22] and Nickel and Kiela [42].
  • domain assumption WordNet nouns, Amazon product categories, and MovieLens movies contain latent hierarchies that matter for the downstream tasks
    The entire evaluation rests on these datasets being hierarchical; Section 4 motivates this but does not quantify the hierarchy.
  • ad hoc to paper A Euclidean norm between Poincaré-ball points is a sufficient training signal for hyperbolic codebooks
    Algorithm 1's l_cmt uses Euclidean differences while nearest-neighbor selection uses hyperbolic distance; the paper never justifies this mismatch.
  • ad hoc to paper A fixed curvature c, never reported, is appropriate for all four datasets
    No curvature sweep or sensitivity analysis is provided; the claim that hyperbolic geometry is the cause of improvement depends on c being appropriate.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hyperbolic Residual Quantization: Discrete Representations for Data with Latent Hierarchies." pith.science (2026). https://pith.science/paper/XH4L466F

@misc{pith2026250512404,
  author       = {Pith},
  title        = {Pith review of: Hyperbolic Residual Quantization: Discrete Representations for Data with Latent Hierarchies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XH4L466F}},
  note         = {Machine review of arXiv:2505.12404}
}
abstract

Hierarchical data arise in countless domains, from biological taxonomies and organizational charts to legal codes and knowledge graphs. Residual Quantization (RQ) is widely used to generate discrete, multitoken representations for such data by iteratively quantizing residuals in a multilevel codebook. However, its reliance on Euclidean geometry can introduce fundamental mismatches that hinder modeling of hierarchical branching, necessary for faithful representation of hierarchical data. In this work, we propose Hyperbolic Residual Quantization (HRQ), which embeds data natively in a hyperbolic manifold and performs residual quantization using hyperbolic operations and distance metrics. By adapting the embedding network, residual computation, and distance metric to hyperbolic geometry, HRQ imparts an inductive bias that aligns naturally with hierarchical branching. We claim that HRQ in comparison to RQ can generate more useful for downstream tasks discrete hierarchical representations for data with latent hierarchies. We evaluate HRQ on two tasks: supervised hierarchy modeling using WordNet hypernym trees, where the model is supervised to learn the latent hierarchy - and hierarchy discovery, where, while latent hierarchy exists in the data, the model is not directly trained or evaluated on a task related to the hierarchy. Across both scenarios, HRQ hierarchical tokens yield better performance on downstream tasks compared to Euclidean RQ with gains of up to $20\%$ for the hierarchy modeling task. Our results demonstrate that integrating hyperbolic geometry into discrete representation learning substantially enhances the ability to capture latent hierarchies.

Figures

Figures reproduced from arXiv: 2505.12404 by the authors.

Figure 1
Figure 1. Recall@10 of the hypernym genera￾tion based on tokens generated by HRQ vs tokens generated by RQ. HRQ consistently outperforms RQ. Furthermore, HRQ sustains consistent scores across different dimensionalities of the embedding. Residual Quantization Variational Autoencoders (RQ-VAE) [32, 60] leverage these benefits by creating semantic hierarchical discrete represen￾tations through a multilevel quantization pro￾cess … view at source ↗
Figure 2
Figure 2. Visualization of the tangent space and related operations. Exponential map expc x maps from the tangent space attached at x to the mani￾fold and logarithmic map logc x maps from the man￾ifold to the tangent space attached at point x. Hyperbolic geometry operates on manifolds with constant negative Gaussian curvature. A fundamental characteristic of hyperbolic geome￾try is its exponential spatial expansion relative t… view at source ↗
Figure 3
Figure 3. HRQ-VAE visualized. In the image HRQ-VAE quantizes given vector [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The embedding space structure induced by RQ-VAE and HRQ-VAE, respectively for a hier￾archical tokens of length 2. The data is represented by coloured dots. Hue of the dot represents first hierarchical token. The shade represents second token. For the RQ-VAE the result …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 46 canonical work pages

  1. [1]

    Claude 3.5 sonnet.https://anthropic.com/claude, 2024

    Anthropic. Claude 3.5 sonnet.https://anthropic.com/claude, 2024

    work page 2024

  2. [2]

    Hyperbolic image segmentation

    Mina Ghadimi Atigh, Julian Schoep, Erman Acar, Nanne Van Noord, and Pascal Mettes. Hyperbolic image segmentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022

    work page 2022

  3. [3]

    Hospedales

    Ivana Balazevic, Carl Allen, and Timothy M. Hospedales. Multi-relational poincaré graph embeddings. InAdvances in Neural Information Processing Sys- tems, 2019. URL https://proceedings.neurips.cc/paper/2019/hash/ f8b932c70d0b2e6bf071729a4fa68dfc-Abstract.html

    work page 2019

  4. [4]

    Riemannian adaptive optimization methods.arXiv preprint arXiv:1810.00760, 2018

    Gary Bécigneul and Octavian-Eugen Ganea. Riemannian adaptive optimization methods.arXiv preprint arXiv:1810.00760, 2018

    arXiv 2018

  5. [5]

    A neural probabilistic language model.Journal of machine learning research, (Feb), 2003

    Yoshua Bengio, Réjean Ducharme, Pascal Vincent, and Christian Jauvin. A neural probabilistic language model.Journal of machine learning research, (Feb), 2003

    work page 2003

  6. [6]

    Stochastic gradient descent on riemannian manifolds.IEEE Transactions on Automatic Control, (9), 2013

    Silvere Bonnabel. Stochastic gradient descent on riemannian manifolds.IEEE Transactions on Automatic Control, (9), 2013

    work page 2013

  7. [7]

    Translating embeddings for modeling multi-relational data

    Antoine Bordes, Nicolas Usunier, Alberto García-Durán, Jason Weston, and Oksana Yakhnenko. Translating embeddings for modeling multi-relational data. InAdvances in Neural Information Processing Systems, 2013. URL https://proceedings.neurips.cc/paper/2013/hash/ 1cecc7a77928ca8133fa24680a88d2f9-Abstract.html

    work page 2013

  8. [8]

    Hyperbolic geometry

    James W Cannon, William J Floyd, Richard Kenyon, Walter R Parry, et al. Hyperbolic geometry. Number 59-115. Citeseer, 1997

    work page 1997

Show all 63 references
  1. [9]

    Hyperbolic deep reinforcement learning.arXiv preprint arXiv:2210.01542, 2022

    Edoardo Cetin, Benjamin Chamberlain, Michael Bronstein, and Jonathan J Hunt. Hyperbolic deep reinforcement learning.arXiv preprint arXiv:2210.01542, 2022. 10

    2022 arXiv

  2. [10]

    Scalable hyperbolic recommender systems.arXiv preprint arXiv:1902.08648, 2019

    Benjamin Paul Chamberlain, Stephen R Hardwick, David R Wardrope, Fabon Dzogang, Fabio Daolio, and Saúl Vargas. Scalable hyperbolic recommender systems.arXiv preprint arXiv:1902.08648, 2019

    1902 arXiv

  3. [11]

    From trees to continuous embeddings and back: Hyperbolic hierarchical clustering.Advances in Neural Information Processing Systems, 2020

    Ines Chami, Albert Gu, Vaggos Chatziafratis, and Christopher Ré. From trees to continuous embeddings and back: Hyperbolic hierarchical clustering.Advances in Neural Information Processing Systems, 2020

    2020

  4. [12]

    Low- dimensional hyperbolic knowledge graph embeddings

    Ines Chami, Adva Wolf, Da-Cheng Juan, Frederic Sala, Sujith Ravi, and Christopher Ré. Low- dimensional hyperbolic knowledge graph embeddings. InProceedings of the 58th Annual Meeting of the Association for Computational (ACL), 2020. URL https://doi.org/10. 18653/v1/2020.acl-main.617

    2020

  5. [13]

    MIT press, 1969

    Alfred D Chandler Jr.Strategy and structure: Chapters in the history of the American industrial enterprise. MIT press, 1969

    1969

  6. [14]

    Modeling scale-free graphs with hyperbolic geometry for knowledge-aware recommendation

    Yankai Chen, Menglin Yang, Yingxue Zhang, Mengchen Zhao, Ziqiao Meng, Jianye Hao, and Irwin King. Modeling scale-free graphs with hyperbolic geometry for knowledge-aware recommendation. InProceedings of the fifteenth ACM international conference on web search and data mining, 2022

    2022

  7. [15]

    Taming transformers for high-resolution image synthesis

    Patrick Esser, Robin Rombach, and Bjorn Ommer. Taming transformers for high-resolution image synthesis. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021

    2021

  8. [16]

    Hyperbolic neural networks.Advances in neural information processing systems, 31, 2018

    Octavian Ganea, Gary Bécigneul, and Thomas Hofmann. Hyperbolic neural networks.Advances in neural information processing systems, 31, 2018

    2018

  9. [17]

    Hyperbolic entailment cones for learning hierarchical embeddings

    Octavian Ganea, Gary Bécigneul, and Thomas Hofmann. Hyperbolic entailment cones for learning hierarchical embeddings. InInternational Conference on Machine Learning. PMLR, 2018

    2018

  10. [18]

    Hyperbolic entailment cones for learning hierarchical embeddings

    Octavian-Eugen Ganea, Gary Bécigneul, and Thomas Hofmann. Hyperbolic entailment cones for learning hierarchical embeddings. InProceedings of the 35th International Conference on Machine Learning (ICML), Proceedings of Machine Learning Research. PMLR, 2018. URL http://proceedin...

    2018

  11. [19]

    Generative adversarial networks.Communications of the ACM, (11), 2020

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial networks.Communications of the ACM, (11), 2020

    2020

  12. [20]

    Hypervq: Mlr-based vector quantiza- tion in hyperbolic space.arXiv preprint arXiv:2403.13015, 2024

    Nabarun Goswami, Yusuke Mukuta, and Tatsuya Harada. Hypervq: Mlr-based vector quantiza- tion in hyperbolic space.arXiv preprint arXiv:2403.13015, 2024

    2024 arXiv

  13. [21]

    Vector quantization.IEEE Assp Magazine, (2), 1984

    Robert Gray. Vector quantization.IEEE Assp Magazine, (2), 1984

    1984

  14. [22]

    Hyperbolic groups.Essays in Group Theory, 1987

    Mikhael Gromov. Hyperbolic groups.Essays in Group Theory, 1987

    1987

  15. [23]

    On the tree-likeness of hyperbolic spaces

    Matthias Hamann. On the tree-likeness of hyperbolic spaces. InMathematical proceedings of the cambridge philosophical society, number 2. Cambridge University Press, 2018

    2018

  16. [24]

    The movielens datasets: History and context.ACM Transactions on Interactive Intelligent Systems (TIIS), (4), 2015

    F Maxwell Harper and Joseph A Konstan. The movielens datasets: History and context.ACM Transactions on Interactive Intelligent Systems (TIIS), (4), 2015

    2015

  17. [25]

    Denoising diffusion probabilistic models.Advances in Neural Information Processing Systems, 2020

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models.Advances in Neural Information Processing Systems, 2020

    2020

  18. [26]

    How to index item ids for recommendation foundation models

    Wenyue Hua, Shuyuan Xu, Yingqiang Ge, and Yongfeng Zhang. How to index item ids for recommendation foundation models. SIGIR-AP ’23. Association for Computing Machinery, 2023

    2023

  19. [27]

    Routledge, 2013

    Barbel Inhelder and Jean Piaget.The early growth of logic in the child: Classification and seriation. Routledge, 2013

    2013

  20. [28]

    Self-attentive sequential recommendation

    Wang-Cheng Kang and Julian McAuley. Self-attentive sequential recommendation. In2018 IEEE international conference on data mining (ICDM). IEEE, 2018

    2018

  21. [29]

    Hyperbolic image embeddings

    Valentin Khrulkov, Leyla Mirvakhabova, Evgeniya Ustinova, Ivan Oseledets, and Victor Lempit- sky. Hyperbolic image embeddings. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020

    2020

  22. [30]

    Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980, 2014

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980, 2014. 11

    2014 arXiv

  23. [31]

    Hyperbolic geometry of complex networks.Physical Review E—Statistical, Nonlinear, and Soft Matter Physics, (3), 2010

    Dmitri Krioukov, Fragkiskos Papadopoulos, Maksim Kitsak, Amin Vahdat, and Marián Boguná. Hyperbolic geometry of complex networks.Physical Review E—Statistical, Nonlinear, and Soft Matter Physics, (3), 2010

    2010

  24. [32]

    Autoregressive image generation using residual quantization

    Doyup Lee, Chiheon Kim, Saehoon Kim, Minsu Cho, and Wook-Shin Han. Autoregressive image generation using residual quantization. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022

    2022

  25. [33]

    Fully hyperbolic rotation for knowledge graph embedding

    Qiuyu Liang, Weihua Wang, Feilong Bao, and Guanglai Gao. Fully hyperbolic rotation for knowledge graph embedding. InProceedings of the 27th European Conference on Artificial Intelligence (ECAI 2024), Frontiers in Artificial Intelligence and Applications. IOS Press, 2024

    2024

  26. [34]

    Fully hyperbolic rotation for knowledge graph embedding

    Qiuyu Liang, Weihua Wang, Feilong Bao, and Guanglai Gao. Fully hyperbolic rotation for knowledge graph embedding. InECAI 2024 - 27th European Conference on Artificial Intelligence, 19-24 October 2024, Santiago de Compostela, Spain - Including 13th Con- ference on Prestigious A...

    2024 doi

  27. [35]

    Hyperexpan: Taxonomy expansion with hyperbolic representation learning.arXiv preprint arXiv:2109.10500, 2021

    Mingyu Derek Ma, Muhao Chen, Te-Lin Wu, and Nanyun Peng. Hyperexpan: Taxonomy expansion with hyperbolic representation learning.arXiv preprint arXiv:2109.10500, 2021

    2021 arXiv

  28. [36]

    The role of systematics in biology: The study of all aspects of the diversity of life is one of the most important concerns in biology.Science, (3815), 1968

    Ernst Mayr. The role of systematics in biology: The study of all aspects of the diversity of life is one of the most important concerns in biology.Science, (3815), 1968

    1968

  29. [37]

    Image-based recommendations on styles and substitutes

    Julian McAuley, Christopher Targett, Qinfeng Shi, and Anton Van Den Hengel. Image-based recommendations on styles and substitutes. InProceedings of the 38th international ACM SIGIR conference on research and development in information retrieval, 2015

    2015

  30. [38]

    A fast file system for unix.ACM Transactions on Computer Systems (TOCS), (3), 1984

    Marshall K McKusick, William N Joy, Samuel J Leffler, and Robert S Fabry. A fast file system for unix.ACM Transactions on Computer Systems (TOCS), (3), 1984

    1984

  31. [39]

    Efficient estimation of word representations in vector space.arXiv preprint arXiv:1301.3781, 2013

    Tomas Mikolov. Efficient estimation of word representations in vector space.arXiv preprint arXiv:1301.3781, 2013

    2013 arXiv

  32. [40]

    George A. Miller. Wordnet: A lexical database for english.Communications of the ACM, (11),

  33. [41]

    Oseledets, and Alexander Tuzhilin

    Leyla Mirvakhabova, Evgeny Frolov, Valentin Khrulkov, Ivan V . Oseledets, and Alexander Tuzhilin. Performance of hyperbolic geometry models on top-n recommendation tasks. In RecSys 2020: Fourteenth ACM Conference on Recommender Systems, Virtual Event, Brazil, September 22-26, ...

    2020

  34. [42]

    Poincaré embeddings for learning hierarchical represen- tations.Advances in Neural Information Processing Systems, 2017

    Maximillian Nickel and Douwe Kiela. Poincaré embeddings for learning hierarchical represen- tations.Advances in Neural Information Processing Systems, 2017

    2017

  35. [43]

    Learning continuous hierarchies in the lorentz model of hyperbolic geometry

    Maximillian Nickel and Douwe Kiela. Learning continuous hierarchies in the lorentz model of hyperbolic geometry. InInternational conference on machine learning. PMLR, 2018

    2018

  36. [44]

    Petrov and Craig Macdonald

    Aleksandr V . Petrov and Craig Macdonald. Generative sequential recommendation with gptrec,

  37. [45]

    Exploring the limits of transfer learning with a unified text-to-text transformer.Journal of machine learning research, (140), 2020

    Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J Liu. Exploring the limits of transfer learning with a unified text-to-text transformer.Journal of machine learning research, (140), 2020

    2020

  38. [46]

    Tran, Jonah Samost, Maciej Kula, Ed H

    Shashank Rajput, Nikhil Mehta, Anima Singh, Raghunandan Hulikal Keshavan, Trung Vu, Lukasz Heldt, Lichan Hong, Yi Tay, Vinh Q. Tran, Jonah Samost, Maciej Kula, Ed H. Chi, and Maheswaran Sathiamoorthy. Recommender systems with generative retrieval. 2024. URL https://openreview....

    2024

  39. [47]

    Rumelhart, Geoffrey E

    David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams. Learning representations by back-propagating errors.Nature, 1986

    1986

  40. [48]

    Representation tradeoffs for hyperbolic embeddings

    Frederic Sala, Chris De Sa, Albert Gu, and Christopher Ré. Representation tradeoffs for hyperbolic embeddings. InInternational conference on machine learning. PMLR, 2018

    2018

  41. [49]

    Schmidhuber

    J. Schmidhuber. A possibility for implementing curiosity and boredom in model-building neural controllers. In J. A. Meyer and S. W. Wilson, editors,Proc. of the International Conference on Simulation of Adaptive Behavior: From Animals to Animats, pages 222–227. MIT Press/Bradf...

    1991

  42. [50]

    Mpnet: Masked and permuted pre-training for language understanding.Advances in Neural Information Processing Systems, 2020

    Kaitao Song, Xu Tan, Tao Qin, Jianfeng Lu, and Tie-Yan Liu. Mpnet: Masked and permuted pre-training for language understanding.Advances in Neural Information Processing Systems, 2020

    2020

  43. [51]

    Hgcf: Hy- perbolic graph convolution networks for collaborative filtering

    Jianing Sun, Zhaoyue Cheng, Saba Zuberi, Felipe Pérez, and Maksims V olkovs. Hgcf: Hy- perbolic graph convolution networks for collaborative filtering. InProceedings of the Web Conference 2021, 2021

    2021

  44. [52]

    World Scientific, 2008

    Abraham Albert Ungar.Analytic hyperbolic geometry and Albert Einstein’s special theory of relativity. World Scientific, 2008

    2008

  45. [53]

    Neural dis- crete representation learning

    Aäron van den Oord, Oriol Vinyals, and Koray Kavukcuoglu. Neural dis- crete representation learning. InAdvances in Neural Information Processing Systems, 2017. URL https://proceedings.neurips.cc/paper/2017/hash/ 7a98af17e63a0ac09ce2e96d03992fbc-Abstract.html

    2017

  46. [54]

    Neural discrete representation learning

    Aaron Van Den Oord, Oriol Vinyals, et al. Neural discrete representation learning. 2017

    2017

  47. [55]

    Gomez, Łukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need.Advances in Neural Information Processing Systems, 2017

    2017

  48. [56]

    Learning phylogenetic trees as hyperbolic point configurations.CoRR, abs/2104.11430, 2021

    Benjamin Wilson. Learning phylogenetic trees as hyperbolic point configurations.CoRR, abs/2104.11430, 2021. URLhttps://arxiv.org/abs/2104.11430

    2021 arXiv

  49. [57]

    Hifi-codec: Group-residual vector quantization for high fidelity audio codec

    Dongchao Yang, Songxiang Liu, Rongjie Huang, Jinchuan Tian, Chao Weng, and Yuexian Zou. Hifi-codec: Group-residual vector quantization for high fidelity audio codec. 2023

    2023

  50. [58]

    Hypformer: Exploring efficient hyperbolic transformer fully in hyperbolic space.arXiv preprint arXiv:2407.01290, 2024

    Menglin Yang, Harshit Verma, Delvin Ce Zhang, Jiahong Liu, Irwin King, and Rex Ying. Hypformer: Exploring efficient hyperbolic transformer fully in hyperbolic space.arXiv preprint arXiv:2407.01290, 2024

    2024 arXiv

  51. [59]

    Vector-quantized image modeling with improved vqgan.arXiv preprint arXiv:2110.04627, 2021

    Jiahui Yu, Xin Li, Jing Yu Koh, Han Zhang, Ruoming Pang, James Qin, Alexander Ku, Yuanzhong Xu, Jason Baldridge, and Yonghui Wu. Vector-quantized image modeling with improved vqgan.arXiv preprint arXiv:2110.04627, 2021

    2021 arXiv

  52. [60]

    Soundstream: An end-to-end neural audio codec

    Neil Zeghidour, Alejandro Luebs, Ahmed Omran, Jan Skoglund, and Marco Tagliasacchi. Soundstream: An end-to-end neural audio codec. IEEE, 2021

    2021

  53. [61]

    S3-rec: Self-supervised learning for sequential recommendation with mutual information maximization

    Kun Zhou, Hui Wang, Wayne Xin Zhao, Yutao Zhu, Sirui Wang, Fuzheng Zhang, Zhongyuan Wang, and Ji-Rong Wen. S3-rec: Self-supervised learning for sequential recommendation with mutual information maximization. InProceedings of the 29th ACM international conference on information...

    2020

  54. [1995]

    doi: 10.1145/219717.219748

  55. [2023]

    URLhttps://arxiv.org/abs/2306.11114

    arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.