arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.00333cs.LG
收藏

通过注意力引导的特征学习实现大语言模型的高效准确操控

Efficient and accurate steering of Large Language Models through attention-guided feature learning

MIT(麻省理工学院);Broad Institute of MIT and Harvard(麻省理工学院-哈佛大学Broad研究所)

Parmida Davarmanesh, Ashia Wilson, Adityanarayanan Radhakrishnan

AI总结:

本文提出注意力引导的特征学习框架,通过解决三个核心挑战,显著提升了大语言模型的操控效率和效果,成功将可操控概念数量几乎翻倍。

摘要:

操控,即直接操纵内部激活以引导LLM响应朝特定语义概念方向发展,正成为理解语义概念如何存储在LLM中的有前途的途径,并推动LLM能力的提升。然而,现有操控方法却相当脆弱,看似不可操控的概念基于提取概念相关特征时算法选择的细微变化而完全变得可操控。在本工作中,我们引入了一个注意力引导的操控框架,克服了与操控相关的三个核心挑战:(1)自动选择提取概念相关特征的相关token嵌入;(2)考虑LLM激活中概念相关特征的异质性;(3)识别对操控最相关的层。在包含512个语义概念的操控基准测试中,我们的框架在各种模型架构和大小(高达700亿参数模型)上显著提高了操控效果(成功操控的概念数量几乎翻倍)。此外,我们利用该框架揭示了LLM各层中概念特定特征的分布情况。总体而言,我们的框架为开发高效、高度可扩展的行业级LLM微调算法开辟了进一步的途径。

英文摘要:

Steering, or direct manipulation of internal activations to guide LLM responses toward specific semantic concepts, is emerging as a promising avenue for both understanding how semantic concepts are stored within LLMs and advancing LLM capabilities. Yet, existing steering methods are remarkably brittle, with seemingly non-steerable concepts becoming completely steerable based on subtle algorithmic choices in how concept-related features are extracted. In this work, we introduce an attention-guided steering framework that overcomes three core challenges associated with steering: (1) automatic selection of relevant token embeddings for extracting concept-related features; (2) accounting for heterogeneity of concept-related features across LLM activations; and (3) identification of layers most relevant for steering. Across a steering benchmark of 512 semantic concepts, our framework substantially improved steering over previous state-of-the-art (nearly doubling the number of successfully steered concepts) across model architectures and sizes (up to 70 billion parameter models). Furthermore, we use our framework to shed light on the distribution of concept-specific features across LLM layers. Overall, our framework opens further avenues for developing efficient, highly-scalable fine-tuning algorithms for industry-scale LLMs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑