NUWA 风险评测 · 研究方法
Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers
NUWA 研究解读 · 独立复核待完成
NUWA Research Note
研究问题研究设计与假设来源发现NUWA 解读局限与相反证据研究与决策意义机构与角色原始资源模型辅助整理 · 选读原始全文或官方文章中的相关章节；尚未完成逐章节审计。
研究速读
NUWA 解读：激活问答可复用接口，但回答仍是模型生成解释。
研究问题
语言模型接口能否跨审计任务解释激活？
研究设计与假设
预言器接收目标激活与自然语言问题，用系统提示、分类及自监督上下文预测训练。
来源发现
作者文章报告泛化至训练未含的秘密信息与微调目标。
四模型族消融中更多样监督提高正确率。
NUWA 解读
NUWA 解读：激活问答可复用接口，但回答仍是模型生成解释。
局限与相反证据
简短作者介绍未确立通用跨模型效度、抗对抗能力或信念访问保证。
研究与决策意义
模型建议：每项下游审计以保留真值验证，并保留目标与预言器身份。
精确来源位置
- Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers · AbstractSelected original resource https://truthful.ai/papers/activation-oracles/; sha256:e57df5c8ebc075889e7fd15723a7f00ff4c789a4d6be58f4899631bb9b50e6e2; archived source capture; exact capture time not independently established
- Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers · MethodSelected original resource https://truthful.ai/papers/activation-oracles/; sha256:e57df5c8ebc075889e7fd15723a7f00ff4c789a4d6be58f4899631bb9b50e6e2; archived source capture; exact capture time not independently established
- Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers · Activation Oracles generalize to downstream auditing tasksSelected original resource https://truthful.ai/papers/activation-oracles/; sha256:e57df5c8ebc075889e7fd15723a7f00ff4c789a4d6be58f4899631bb9b50e6e2; archived source capture; exact capture time not independently established
- Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers · How does Activation Oracle training scale?Selected original resource https://truthful.ai/papers/activation-oracles/; sha256:e57df5c8ebc075889e7fd15723a7f00ff4c789a4d6be58f4899631bb9b50e6e2; archived source capture; exact capture time not independently established
来源核查、数据提取、编辑批准与实验复现分别记录。 方法说明 →
机构与角色
作者 (11)
Adam Karvonen · James Chua · Clément Dumas · Kit Fraser-Taliente · Subhash Kantamneni · Julian Minder · Euan Ong · Arnab Sen Sharma · Daniel Wen · Owain Evans · Samuel Marks
引用与复用
引用研究发现时请引用原始出版物；引用 NUWA 的整理与解释时，单独引用本站记录及对应快照。
原始出版物 · BibTeX
@misc{activation-oracles,
title = {Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers},
url = {https://truthful.ai/papers/activation-oracles/},
author = {Adam Karvonen and James Chua and Clément Dumas and Kit Fraser-Taliente and Subhash Kantamneni and Julian Minder and Euan Ong and Arnab Sen Sharma and Daniel Wen and Owain Evans and Samuel Marks},
year = {2025},
date = {2025-12-19},
version = {Source capture 2026-10-25},
note = {Source capture 2026-10-25}
}
NUWA 记录 · BibTeX
@misc{activation-oracles-nuwa,
title = {Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers — NUWA research record},
url = {https://whitzard.tech/nuwa/whitzard-index/work/activation-oracles/},
version = {2026-10-29.1},
organization = {NUWA Frontier AI Safety Lab},
howpublished = {Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--activation-oracles.json},
note = {NUWA research note. Source-reading scope and independent review status are stated on the record. Snapshot 2026-10-29.1. Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--activation-oracles.json.}
}
证据快照 2026-10-29.1 · 索引更新 · RSS
参考文献
- [1] Adam Karvonen, James Chua 等. Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers. 2025-12-19 ; Official publication masthead and article byline; Abstract; Method