Sakana Fugu Technical Report
- lab arXiv
- lab arXivLabs
- person Stefan Nielsen
Researchers have developed two new AI models, Sakana Fugu and LemonHarness, designed to improve performance in complex tasks by harnessing the capabilities of multiple large language models (LLMs).
Sakana Fugu is a family of orchestrator models that combine the specializations of various LLMs into a collectively intelligent system. Fugu models are trained to understand user queries and dynamically devise agentic scaffolds to solve them, achieving state-of-the-art results across challenging tasks [1]. Two Fugu models have been released: Fugu, which balances performance with latency for everyday use, and Fugu-Ultra, which prioritizes answer quality on the hardest problems. Meanwhile, LemonHarness is a framework that constrains state-changing operations within a clearly defined workspace and introduces a reusable rule knowledge base. It also adds a time-aware execution mechanism, exposing elapsed and remaining budget to the model [2]. LemonHarness was tested on Terminal-Bench 2.0, achieving 84.49% accuracy over 445 trials with GPT-5.3-CodeX, and 86.52% accuracy with GPT-5.5 [2].
benchmarkapplicationmodel-releaseresearch-paper
Background sources we checked (10)
- arxiv.org ↗ The capabilities of frontier Large Language Models (LLMs) continue to advance, with different providers increasingly specializing in distinct domains. This raises a natural next objective: how to combine the individual specializations of various LLMs into a collectively intellige…
- arxiv.org ↗ # Sakana Fugu Technical Report ... The capabilities of frontier Large Language Models (LLMs) continue to advance, with different providers increasingly specializing in distinct domains. This raises a natural next objective: how to combine the individual specializations of various…
- arxiv.org ↗ # Sakana Fugu Technical Report ... The capabilities of frontier Large Language Models (LLMs) continue to advance, with different providers increasingly specializing in distinct domains. This raises a natural next objective: how to combine the individual specializations of various…
- huggingface.co ↗ Title: Sakana Fugu Technical Report ... The capabilities of frontier Large Language Models (LLMs) continue to advance, with different providers increasingly specializing in distinct domains. This raises a natural next objective: how to combine the individual specializations of va…
- info.arxiv.org ↗ arXiv Labs - arXiv info | arXiv e-print repository Skip to content # arXiv Labs Attention arXiv Users: arXiv Labs is pausing new proposals ## What are arXiv Labs? arXiv Labs are a way for the community to contribute new, useful features to arXiv. These integrations are avail…
- blog.arxiv.org ↗ arXivLabs: a space for community innovation – arXiv blog arXiv has launched a new, formalized framework enabling innovative collaborations with individuals and organizations. “Members of our community want to contribute tools that enhance the arXiv experience, and we val…
- info.arxiv.org ↗ arXivLabs: Showcase - arXiv info | arXiv e-print repository ... # arXivLabs: Showcase ... arXiv is surrounded by a community of researchers and developers working at the cutting edge of information science and technology. ... While the arXiv team is focused on our core mission—pr…
- en.wikipedia.org ↗ arXiv (pronounced as "archive"—the X represents the Greek letter chi ⟨χ⟩) is an open-access repository of electronic preprints and postprints (known as e-prints) approved for posting after moderation, but not peer reviewed. It consists of scientific papers in the fields of mathem…
- en.wikipedia.org ↗ 14 (fourteen) is the natural number following 13 and preceding 15.…
- en.wikipedia.org ↗ A large language model (LLM) is a type of machine learning model designed for natural language processing tasks such as language generation. LLMs are language models with many parameters, and are trained with self-supervised learning on a vast amount of text.…