牛哥精选 · 所有

📋 全部 🤖 AI·大模型 ⚡ 效率工具 📝 深度技术 🚀 产品观察 💰 商业科技 🔓 开源项目 🎨 设计创意 📖 阅读推荐 🏷 资源合集 🌱 成长效率

📝 深度技术 arXiv 机器学习 2026-06-15

Generalizing GNNs with Tokenized Mixture of Experts

KDD 2026论文提出Tokenized MoE框架，用专家混合机制突破GNN泛化瓶颈

arXiv:2602.09258v2 Announce Type: replace Abstract: Deployed graph neural networks (GNNs) are frozen at deployment yet must fit clean data, generalize…

gnn mixture of 图神经网络泛化 tokenized

📝 深度技术 arXiv NLP 2026-06-05

CoMoL: Efficient Mixture of LoRA Experts via Dynamic Core Space Merging

提出动态核心空间合并方法，大幅降低混合LoRA专家模型的内存占用与计算开销

arXiv:2603.00573v2 Announce Type: replace Abstract: Large language models (LLMs) achieve remarkable performance on diverse downstream and domain-speci…

comol mixture of dynamic co 模型压缩参数高效微调

📝 深度技术 arXiv 机器学习 2026-05-28

Attention Sink Forges Native MoE in Attention Layers: Sink-Aware Training to Address Head Collapse

利用注意力汇合现象，在注意力层内实现原生MoE训练，有效解决头部坍塌问题。

arXiv:2602.01203v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) often assign disproportionate attention to the first token, a p…

attention mixture of head colla 模型训练深度学习

📝 深度技术 arXiv NLP 2026-05-20

Mixture of Experts for Low-Resource LLMs

揭秘低资源语言大模型中的专家路由：对比Transformer与Mamba混合架构的MoE表现。

arXiv:2605.17598v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) architectures enable efficient model scaling, yet expert routing behavior acr…

moe 低资源语言专家路由 transforme mamba

📅 日期

2026-05-20 2026-05-19

🐂 牛哥精选

Generalizing GNNs with Tokenized Mixture of Experts

CoMoL: Efficient Mixture of LoRA Experts via Dynamic Core Space Merging

Attention Sink Forges Native MoE in Attention Layers: Sink-Aware Training to Address Head Collapse

Mixture of Experts for Low-Resource LLMs

📅 日期