---
title: "面向AI治理的后训练自适应技术六维分类法"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.06246v1"
permalink: "https://feed.astralor.com/items/itm_ce00c7b3c386b5fe6eda188e"
published: "2026-08-06T16:32:26.000Z"
score: 75
channel: "ai"
tags: ["AI治理","分类法","后训练自适应","微调","检索增强"]
---

# 面向AI治理的后训练自适应技术六维分类法

## AI 点评

> 这篇综述最吸引我的是它试图终结'微调'等术语的滥用。对于需要审计模型变更的团队来说，这套分类法可以成为内部文档的模板。此外，它指出的开放挑战，比如推理时自适应和多模态适应，正是下一波研究热点，值得持续关注。

## 速览

后训练自适应已成为现代机器学习核心，涵盖重新训练、微调、参数高效自适应、对齐、检索增强、模型编辑、遗忘、校准和多模态指令微调等。然而文献碎片化，难以比较方法或描述模型修改。本综述提出六维分类法，按机制、目标、数据需求、持久性、结构范围和模型类型组织，区分了微调、检索增强与提示等易混淆概念，并梳理了从传统机器学习到多模态大语言模型的演化路径，以及技术间的继承、取代、混合和分层部署关系。该词汇体系可支持技术文档、模型变更追踪与治理分析，并在评估、可复现性、推理时自适应、遗忘、多模态适应等方面提出开放挑战。

## 关键点

- 提出六维分类法，统一后训练自适应技术术语。
- 区分微调、检索增强与提示等易混概念。
- 梳理技术演进及继承、取代、混合和分层部署关系。
- 指出评估、可复现性、多模态适应等开放挑战。

## 全文

Post-training adaptation has become central to modern machine learning practice and includes techniques such as retraining, fine-tuning, parameter-efficient adaptation, alignment, retrieval augmentation, model editing, unlearning, calibration, and Multimodal Instruction Tuning. However, the literature remains fragmented across technique families, model classes, and deployment contexts, making it difficult to compare methods or describe how a trained model has been modified. This survey synthesizes the post-training adaptation literature and introduces a six-dimensional taxonomy organized by mechanism, goal, data requirement, persistence, structural scope, and model type. The taxonomy distinguishes commonly conflated terms such as fine-tuning, retrieval augmentation, and prompting, and shows how adaptation strategies evolve from traditional machine learning through deep learning, foundation models, large language models, and multimodal large language models. It also maps relationships among techniques, including inheritance, supersession, hybridization, and layered deployment stacks. The resulting vocabulary can support technical documentation, model-change tracking, and governance analysis. The survey concludes by identifying open challenges in evaluation, reproducibility, persistent inference-time adaptation, unlearning, multimodal adaptation, and governance-aware post-training workflows.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_ce00c7b3c386b5fe6eda188e