HEBATRON: A Hebrew-Specialized Open-Weight Mixture-of-Experts Language Model
Researchers present Hebatron, an open-weight Hebrew LLM built on NVIDIA Nemotron-3 MoE architecture. It achieves 73.8% Hebrew reasoning accuracy with only 3B active parameters per forward pass, outperforming prior models and rivaling larger models like Gemma-3-27B, while providing 9x inference throughput and 65k token context.
Article intelligence
Key points
- First open-weight Hebrew-specialized MoE model with native long-context support.
- Employs a three-phase easy-to-hard curriculum with anti-forgetting anchoring and fine-tuning on 2M bilingual samples.
- Achieves 73.8% Hebrew reasoning average, outperforming DictaLM-3.0-24B-Thinking (68.9%).
- Activating only 3B parameters per forward pass from a 30B model enables ~9x inference throughput.
Why it matters
This matters because first open-weight Hebrew-specialized MoE model with native long-context support.
Technical impact
May affect model selection, inference cost, product capability, and evaluation benchmarks.
[2605.11255] HEBATRON: A Hebrew-Specialized Open-Weight Mixture-of-Experts Language Model
[Submitted on 11 May 2026]
Title:HEBATRON: A Hebrew-Specialized Open-Weight Mixture-of-Experts Language Model
View a PDF of the paper titled HEBATRON: A Hebrew-Specialized Open-Weight Mixture-of-Experts Language Model, by Noam Kayzer and 12 other authors
View PDF HTML (experimental)
Abstract:We present Hebatron, a Hebrew-specialized open-weight large language model built on the NVIDIA Nemotron-3 sparse Mixture-of-Experts architecture. Training employs a three-phase easy-to-hard curriculum with continuous anti-forgetting anchoring, followed by supervised fine-tuning on 2 million bilingual Hebrew--English samples. The curriculum ordering alone yields a 3-point aggregate benchmark gain over the reversed configuration. Hebatron achieves a Hebrew reasoning average of 73.8\%, outperforming DictaLM-3.0-24B-Thinking (68.9\%) and remaining competitive with Gemma-3-27B-IT on GSM8K-HE and Israeli Trivia, while activating only 3B parameters per forward pass across a 30B-parameter model, delivering approximately 9 times higher inference throughput at native context lengths up to 65,536 tokens. To our knowledge, this is the first language-specific adaptation of the Nemotron-3 architecture for any target language, and the first open-weight Hebrew-specialized MoE model with native long-context support. Model weights are released openly to support further research in Hebrew and Semitic-language NLP.
Subjects:
Computation and Language (cs.CL)
Cite as: arXiv:2605.11255 [cs.CL]
(or arXiv:2605.11255v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2605.11255
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Sarel Weinberger [view email] [v1] Mon, 11 May 2026 21:27:53 UTC (257 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled HEBATRON: A Hebrew-Specialized Open-Weight Mixture-of-Experts Language Model, by Noam Kayzer and 12 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.CL
new | recent | 2026-05
Change to browse by:
cs
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
Loading...
Data provided by:
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)