---
title: "忠实于拒绝：神经元选择器的因果审计"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.05355v1"
permalink: "https://feed.astralor.com/items/itm_b6e952d88681c0512ded070d"
published: "2026-07-06T17:33:36.000Z"
score: 79
channel: "ai"
tags: ["可解释性","因果审计","大语言模型","归因方法","模型安全","模型编辑"]
---

# 忠实于拒绝：神经元选择器的因果审计

## AI 点评

> 这篇论文直接点出了归因方法评估中的一个常见盲区：秩稳定性高不一定意味着因果有效。如果你从事模型编辑或安全对齐，这篇文章值得细读，它对现有工具的有效性提出了硬核拷问。

## 速览

本文通过零化单样本神经元行的方式，对五个大型语言模型的归因方法进行直接因果审计。研究发现归因方法在识别可丢弃行方面优于激活和幅度基线，但高度秩稳定的选择器可能因果有效性较低。拒绝功能存在于冗余子空间中，不同归因方法通过几乎不重叠的行集实现拒绝编辑，表明编辑只是充分集的一个实现而非唯一机制。该研究挑战了依赖秩稳定性代理评估选择器有效性的常见做法。

## 关键点

- 归因方法在识别语言模型中可丢弃神经元行方面优于激活和幅度基线
- 高度秩稳定的选择器可能因果有效性较低，直接因果审计能发现代理指标遗漏的失败模式
- 拒绝功能存在于冗余子空间，不同归因方法通过几乎不重叠的行集实现拒绝编辑
- 研究提供了可直接测试归因方法因果重要性的审计框架，对模型编辑和安全对齐有方法论启示

## 全文

Attribution scores increasingly identify which neuron rows of a language model matter for applications such as pruning, interpretability, and editing for safety, yet whether they identify causally important rows is rarely tested directly. We address this with two paired audits built on one-shot neuron-row zeroing. We first audit selectors at the language-modeling level: attribution methods substantially outperform activation and magnitude-based baselines at identifying dispensable rows across five LLMs. We then adapt the same intervention into a behavior test by driving it with a contrastive harmful-versus-benign signal; the attributed rows are sufficient to install refusal on hate and crime while keeping benign over-refusal low and preserving language model fluency, and specific in that layer-matched random controls at the same depths fail. Highly rank-stable selectors can be among the least causally valid. Refusal moreover lives in a redundant subspace, where different attribution methods install it through largely disjoint row sets, so the recovered edit is one realization of a sufficient set rather than a unique mechanism. Together, these findings show that rank-stability proxies miss the kinds of selector failures a direct causal audit can surface.%

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_b6e952d88681c0512ded070d