English(EN) Instella-MoE Technical Report

Instella-MoE：新的开源 MoE 语言模型发布

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-02 04:00

一份新的技术报告介绍 Instella-MoE，一个拥有 160 亿总参数的开源专家混合（MoE）语言模型。该模型在 AMD Instinct GPU 上训练，并采用了门控多头潜在注意力（Gated Multi-head Latent Attention）和 FarSkip-Collective 连接等创新技术，以实现高效的训练和推理。Instella-MoE 在预训练基准测试中表现强劲，优于其他开源模型，并在指令遵循、推理、数学、编码和聊天任务上取得了有竞争力的结果。开发者正在发布模型权重、训练配置、数据混合和代码，以促进可复现的研究。 AI

影响为高效的 MoE 模型和该领域的可复现研究提供了一个全新的、完全开放的基础。

排序理由该集群包含一份详细介绍新开源语言模型发布的技术报告。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 25 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一份详细介绍新开源语言模型发布的技术报告。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: model release, paper, infra; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Jiang Liu, Sudhanshu Ranjan, Prakamya Mishra, Yonatan Dukler, Gowtham Ramesh, Jialian Wu, Ximeng Sun, Wen Xie, Chaojun Hou, Vikram Appia, Zhenyu Gu, Zicheng Liu, Emad Barsoum · 2026-09-02 04:00

Instella-MoE 技术报告

arXiv:2609.00791v1 Announce Type: cross Abstract: In this work, we introduce Instella-MoE, a fully open Mixture-of-Experts (MoE) language model with 16 billion total parameters and 2.8 billion active parameters per token, trained entirely from scratch on AMD Instinct MI300X and M…

本文如何被排名

报道来源 [1]

Instella-MoE 技术报告

相关实体

相关话题