---
title: "文化常识有表征但无法解码：18个开源大模型中的神话知识溯源"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.02486v1"
permalink: "https://feed.astralor.com/items/itm_7347d842a6ee605ebcdf6f6c"
published: "2026-08-03T16:53:15.000Z"
score: 78
channel: "ai"
tags: ["可解释性","开源大模型","文化偏见","激活修补","线性探针","跨文化神话知识"]
---

# 文化常识有表征但无法解码：18个开源大模型中的神话知识溯源

## AI 点评

> 这篇论文的价值在于把“模型有知识但答不对”从经验观察变成可定位的计算事实。对那些部署多语言模型的团队来说，它提醒我们输出层才是文化偏见真正的瓶颈；对研究社区，它给出了一个容易复用的分析框架——你可以在自己的模型上先跑一遍逐实体探针，再决定是改训练还是改解码。

## 速览

论文对8个架构家族的18个开源大模型进行跨文化神话知识探测，发现模型在残差流中能清晰区分不同文化的神话实体，但解码层会将这些文化特有token坍缩为西方主流传统。作者通过线性探针、logit lens、激活修补和输出提取定位失败环节，表明问题出在读出阶段而非表征阶段；同时发现提示语言会门控解码器，跨语言表现彼此解耦。研究发布了逐实体分解框架、引用锚定的跨文化基准及18个模型的预测结果。

## 关键点

- 模型能区分不同文化的神话实体，但在输出时把芬兰、斯拉夫、埃及、中国等较少被代表传统的神话角色坍缩为希腊-罗马或北欧主流对应物。
- 线性探针和激活修补显示残差流中含有文化区分信息，代表问题不在知识表征，而是解码器读出环节。
- 提示语言显著影响解码行为：用目标文化母语提问与用英语提问的失败模式按语言聚类，跨语言之间解耦。
- 资源开放：逐实体探针与输出分解框架、引用锚定跨文化基准、18个模型的全量预测均已发布。

## 全文

Open-source LLMs reliably name Zeus, Jupiter, and Thor, but recover their counterparts in less-represented traditions like Finnish, Slavic, Egyptian, or Chinese mythology far less consistently. We ask where inside the model this cultural default is produced. On a parallel cross-cultural substrate of Thompson-motif entities, we instrument 18 open-source LLMs from 8 architecture families with linear probing, logit lens, activation patching, and output extraction. The residual stream cleanly distinguishes cultures, well above a name-string baseline, yet the decoder collapses culturally-specific tokens onto dominant-tradition ones. The failure is at readout, not at representation. Asking the same question in the target culture's native language versus English produces failures that cluster within language but decouple across language: the decoder is gated on prompt language. We release a per-entity (probe, output) decomposition framework, a citation-anchored cross-cultural ground truth, a within- versus cross-mode correlation test for language-conditioned readout, and per-entity predictions for all 18 models.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_7347d842a6ee605ebcdf6f6c