Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget Control
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914543009005568 |
|---|---|
| author | Taghibakhshi, Ali Cai, Ruisi Muralidharan, Saurav Sreenivas, Sharath Turuvekere Vavre, Aditya Mahabaleshwarkar, Ameya Sunil Kartal, Bilal Liang, Sheldon Chochowski, Marcin Chen, Zijia Bercovich, Akhiad Zilberstein, Ran El-Yaniv, Ran Geifman, Yonatan Korzekwa, Daniel Suhara, Yoshi Olabiyi, Oluwatobi Aithal, Ashwath Tajbakhsh, Nima Molchanov, Pavlo |
| author_facet | Taghibakhshi, Ali Cai, Ruisi Muralidharan, Saurav Sreenivas, Sharath Turuvekere Vavre, Aditya Mahabaleshwarkar, Ameya Sunil Kartal, Bilal Liang, Sheldon Chochowski, Marcin Chen, Zijia Bercovich, Akhiad Zilberstein, Ran El-Yaniv, Ran Geifman, Yonatan Korzekwa, Daniel Suhara, Yoshi Olabiyi, Oluwatobi Aithal, Ashwath Tajbakhsh, Nima Molchanov, Pavlo |
| contents | Training a family of large language models (LLMs), either from scratch or via iterative compression, is prohibitively expensive and inefficient, requiring separate training runs for each model in the family. In this paper, we introduce Star Elastic, a novel LLM post-training method that adds N nested submodels to a given parent reasoning model using the compute of one run (N-fold savings) via a single post-training job. Beyond reducing training costs, Star Elastic also addresses a fundamental limitation of efficient reasoning: the rigidity of static architectures, which forces the allocation of constant resources regardless of token difficulty. By unlocking elastic budget control, Star Elastic enables a novel inference scheme that uses different submodels for each reasoning phase (thinking and answering). Star Elastic supports (1) nesting along the SSM, embedding channel, MoE, and FFN axes, (2) learning nested submodels via an end-to-end trainable router, and (3) curriculum-based knowledge distillation. Building on the Nemotron Elastic framework, we apply Star Elastic to the NVIDIA Nemotron Nano models, with a particular focus on hybrid Mixture-of-Experts (MoE) architectures: from Nemotron Nano v3 (30B/3.6A), we generate 23B (2.8A) and 12B (2.0A) variants with 160B training tokens. All nested models match or outperform independently trained baselines of comparable size and achieve a 360x reduction versus pretraining from scratch and a 7x reduction over state-of-the-art compression. Crucially, elastic budget control advances the accuracy-latency Pareto frontier, achieving up to 16% higher accuracy and 1.9x lower latency via dynamic per-phase model selection. We further extend Star Elastic to quantized regimes via Quantization-Aware Distillation (QAD), producing nested NVFP4 and FP8 elastic checkpoints that preserve zero-shot slicing while delivering smaller deployment footprints. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_07182 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget Control Taghibakhshi, Ali Cai, Ruisi Muralidharan, Saurav Sreenivas, Sharath Turuvekere Vavre, Aditya Mahabaleshwarkar, Ameya Sunil Kartal, Bilal Liang, Sheldon Chochowski, Marcin Chen, Zijia Bercovich, Akhiad Zilberstein, Ran El-Yaniv, Ran Geifman, Yonatan Korzekwa, Daniel Suhara, Yoshi Olabiyi, Oluwatobi Aithal, Ashwath Tajbakhsh, Nima Molchanov, Pavlo Machine Learning Training a family of large language models (LLMs), either from scratch or via iterative compression, is prohibitively expensive and inefficient, requiring separate training runs for each model in the family. In this paper, we introduce Star Elastic, a novel LLM post-training method that adds N nested submodels to a given parent reasoning model using the compute of one run (N-fold savings) via a single post-training job. Beyond reducing training costs, Star Elastic also addresses a fundamental limitation of efficient reasoning: the rigidity of static architectures, which forces the allocation of constant resources regardless of token difficulty. By unlocking elastic budget control, Star Elastic enables a novel inference scheme that uses different submodels for each reasoning phase (thinking and answering). Star Elastic supports (1) nesting along the SSM, embedding channel, MoE, and FFN axes, (2) learning nested submodels via an end-to-end trainable router, and (3) curriculum-based knowledge distillation. Building on the Nemotron Elastic framework, we apply Star Elastic to the NVIDIA Nemotron Nano models, with a particular focus on hybrid Mixture-of-Experts (MoE) architectures: from Nemotron Nano v3 (30B/3.6A), we generate 23B (2.8A) and 12B (2.0A) variants with 160B training tokens. All nested models match or outperform independently trained baselines of comparable size and achieve a 360x reduction versus pretraining from scratch and a 7x reduction over state-of-the-art compression. Crucially, elastic budget control advances the accuracy-latency Pareto frontier, achieving up to 16% higher accuracy and 1.9x lower latency via dynamic per-phase model selection. We further extend Star Elastic to quantized regimes via Quantization-Aware Distillation (QAD), producing nested NVFP4 and FP8 elastic checkpoints that preserve zero-shot slicing while delivering smaller deployment footprints. |
| title | Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget Control |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2605.07182 |