---
id: 20261118-T0-12
title: "GRAFT：跨骨干网络复用LoRA，实现持续学习"
title_en: "GRAFT Reuses LoRA Across Backbones for Continual Learning"
url: https://ai.daily.yangsir.net/daily/20261118-T0-12
issue_date: 2026-11-18
publish_date: 2026-11-17T05:00:00.000Z
category: research
source_name: "arXiv cs.LG (ML)"
source_url: https://arxiv.org/abs/2611.18638
---

# GRAFT：跨骨干网络复用LoRA，实现持续学习

新论文提出 GRAFT 方法，解决持续学习中预训练 LoRA 适配器难以跨模型复用的问题。现有做法通常要求适配器与模型共享同一骨干网络，限制了复用范围。GRAFT 实现跨骨干网络的 LoRA 复用，让语言模型和多模态模型在持续学习时能借用其他模型已训练好的适配器。对需要频繁更新模型能力的团队来说，这可以减少重复训练适配器的成本。

## English Version

**GRAFT Reuses LoRA Across Backbones for Continual Learning**

A new paper introduces GRAFT, a method addressing the difficulty of reusing pretrained LoRA adapters across models in continual learning. Existing approaches typically require adapters and models to share the same backbone, limiting reuse. GRAFT enables cross-backbone LoRA reuse, letting language and multimodal models borrow already-trained adapters from other models during continual learning. For teams that frequently update model capabilities, this can reduce the cost of retraining adapters from scratch.

---

**来源**：[arXiv cs.LG (ML)](https://arxiv.org/abs/2611.18638)

**详情页**：https://ai.daily.yangsir.net/daily/20261118-T0-12

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*