Conference paper · 2025 · Open access

HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts

Neil He, Rishabh Anand, Hiren Madhu, Ali Maatouk, Smita Krishnaswamy, Leandros Tassiulas, Menglin Yang, Rex Ying

neural information processing systems, 158268–158299 · Published 1 January 2025 · DOI 10.52202/085713-4774

Open PDFView at the publisherCite this paper

In brief

“HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts” is a 2025 conference paper by Neil He and 7 co-authors, published in neural information processing systems. OpenAlex counts 32 citations as of 18 November 2026. It is open access under a CC BY licence: the full text is free to read at arXiv (Cornell University).

Citations
32
Most cited in
2026 (24)
Field-weighted impact
6.34
References
34
Authors
8

Abstract

Large language models (LLMs) have shown great success in text modeling tasks across domains. However, natural language exhibits inherent semantic hierarchies and nuanced geometric structure, which current LLMs do not capture completely owing to their reliance on Euclidean operations. Recent studies have also shown that not respecting the geometry of token embeddings leads to training instabilities and degradation of generative capabilities. These findings suggest that shifting to non-Euclidean geometries can better align language models with the underlying geometry of text. We thus propose to operate fully in Hyperbolic space, known for its expansive, scale-free, and low-distortion properties. We thus introduce HELM, a family of HypErbolic Large Language Models, offering a geometric rethinking of the Transformer-based LLM that addresses the representational inflexibility, missing set of necessary operations, and poor scalability of existing hyperbolic LMs. We additionally introduce a Mixture-of-Curvature Experts model, HELM-MICE, where each expert operates in a distinct curvature space to encode more fine-grained geometric structure from text, as well as a dense model, HELM-D. For HELM-MICE, we further develop hyperbolic Multi-Head Latent Attention (HMLA) for efficient, reduced-KV-cache training and inference. For both models, we develop essential hyperbolic equivalents of rotary positional encodings and RMS normalization. We are the first to train fully hyperbolic LLMs at billion-parameter scale, and evaluate them on well-known benchmarks such as MMLU and ARC, spanning STEM problem-solving, general knowledge, and commonsense reasoning. Our results show consistent gains from our HELM architectures -- up to 4% -- over popular Euclidean architectures used in LLaMA and DeepSeek, highlighting the efficacy and enhanced reasoning afforded by hyperbolic geometry in large-scale LM pretraining.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

He, N., Anand, R., Madhu, H., Maatouk, A., Krishnaswamy, S., Tassiulas, L., Yang, M., & Ying, R. (2025). HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts. neural information processing systems, 158268–158299. https://doi.org/10.52202/085713-4774

MLA 9

He, Neil, et al. “HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts.” neural information processing systems, 2025, pp. 158268–158299. https://doi.org/10.52202/085713-4774.

Chicago

He, Neil, Rishabh Anand, Hiren Madhu, Ali Maatouk, Smita Krishnaswamy, Leandros Tassiulas, Menglin Yang, and Rex Ying. 2025. “HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts.” neural information processing systems: 158268–158299. https://doi.org/10.52202/085713-4774.

Harvard

He, N. et al. (2025) ‘HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts’, neural information processing systems, pp. 158268–158299. doi:10.52202/085713-4774.

IEEE

N. He et al., “HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts,” neural information processing systems, pp. 158268–158299, 2025, doi: 10.52202/085713-4774.

BibTeX

@inproceedings{he2025helm,
  title     = {{HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts}},
  author    = {He, Neil and Anand, Rishabh and Madhu, Hiren and Maatouk, Ali and Krishnaswamy, Smita and Tassiulas, Leandros and Yang, Menglin and Ying, Rex},
  booktitle = {neural information processing systems},
  year      = {2025},
  pages     = {158268--158299},
  doi       = {10.52202/085713-4774},
  url       = {https://doi.org/10.52202/085713-4774}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Citations by year

YearCitationsShare of the best year
202211
202577
20262424

Its field-weighted citation impact is 6.34: it has been cited 6.34 times as often as the average work of the same field, type and year (1.00 is the world average).

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling

  • hyperbolic space
  • rotary positional encoding
  • non-Euclidean geometry
  • semantic hierarchy
  • commonsense reasoning

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

He, N., Anand, R., Madhu, H., Maatouk, A., Krishnaswamy, S., Tassiulas, L., Yang, M., & Ying, R. (2025). HELM: Hyperbolic Large Language Models via Mixture-of-Curvature Experts. neural information processing systems, 158268–158299. https://doi.org/10.52202/085713-4774

How many times has this paper been cited?

32 times, according to OpenAlex on 18 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.

Who wrote this paper?

Neil He, Rishabh Anand, Hiren Madhu, Ali Maatouk, Smita Krishnaswamy, Leandros Tassiulas, Menglin Yang, Rex Ying.

What is the DOI of this paper?

10.52202/085713-4774, permanently at https://doi.org/10.52202/085713-4774. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 18 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.