English(EN) Monarch-MoE: Structured Experts for Multimodal Continual Instruction Tuning

Monarch-MoE：用于多模态持续指令调优的结构化专家

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-12 04:00

研究人员开发了Monarch-MoE，一种用于多模态持续指令调优的新方法，解决了灾难性遗忘问题。通过使用块对角因子和固定排列来构建LoRA专家，Monarch-MoE实现了块状梯度隔离，这意味着一个块的更新不会影响其他块。该方法在MLLM-DCL和UCIT的评估中分别显著减少了适配器参数（约75%和50%），同时保持了较高的首任务保留率。该系统针对内存有限的硬件进行了优化，能够在单块24 GB RTX 4090上进行微调和推理，运行时长具有竞争力且内存使用量较低。 AI

影响使得在内存受限的硬件上，多模态模型能够进行更高效、更有效的持续学习。

排序理由这是一篇研究论文，详细介绍了一种新的多模态持续指令调优方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 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: 11 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 这是一篇研究论文，详细介绍了一种新的多模态持续指令调优方法。[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: 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: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Xi Zhang, Boyu Diao, Zesen Chen, Ruiqi Liu, Hangda Liu, Yongjun Xu · 2026-10-12 04:00

Monarch-MoE：用于多模态持续指令微调的结构化专家

arXiv:2610.13215v1 Announce Type: new Abstract: Learning new tasks in multimodal continual instruction tuning can cause catastrophic forgetting of previously acquired knowledge. Although multi-expert LoRA distributes learning across experts, updates to a dense low-rank expert can…

本文如何被排名

报道来源 [1]

Monarch-MoE：用于多模态持续指令微调的结构化专家

相关实体

相关话题