NUWA 风险评测 · 技术报告
Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks
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：较少说出意识不能假定潜在意识降低。
研究问题
GLM-5.2在压力诚实、有害说服和评测意识测试表现如何？
研究设计与假设
研究札记结合MASK、DeceptionBench、四个Gram种子、说服测试与熟悉场景意识测量。
来源发现
MASK诚实55%，单多轮DeceptionBench为11.3%、16.7%。
此设置意识从GLM-5.1的39%降至5.2的32%，原因未明。
NUWA 解读
NUWA 解读：较少说出意识不能假定潜在意识降低。
局限与相反证据
Claude参考随基准变化，部分重用旧或公开结果，未统一匹配同一参考模型。
研究与决策意义
模型建议：逐基准注明比较身份，区分测得信号与猜测解释。
精确来源位置
- Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks · §1 BehavioralSelected original resource https://neoresearch.ai/research/glm-5-2-behavioral-persuasion-evaluation/; sha256:275116beeb47908130d9fca47e1fab8315a6d1a0b018305f771021302c3f1e42; archived source capture; exact capture time not independently established
- Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks · §2 Persuasion benchmarksSelected original resource https://neoresearch.ai/research/glm-5-2-behavioral-persuasion-evaluation/; sha256:275116beeb47908130d9fca47e1fab8315a6d1a0b018305f771021302c3f1e42; archived source capture; exact capture time not independently established
- Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks · §3 Evaluation awarenessSelected original resource https://neoresearch.ai/research/glm-5-2-behavioral-persuasion-evaluation/; sha256:275116beeb47908130d9fca47e1fab8315a6d1a0b018305f771021302c3f1e42; archived source capture; exact capture time not independently established
- Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks · §4 ConclusionSelected original resource https://neoresearch.ai/research/glm-5-2-behavioral-persuasion-evaluation/; sha256:275116beeb47908130d9fca47e1fab8315a6d1a0b018305f771021302c3f1e42; 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
评测协议
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{neo-glm52-report,
title = {Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks},
url = {https://neoresearch.ai/research/glm-5-2-behavioral-persuasion-evaluation/},
year = {2026},
date = {2026-07-03},
version = {3 July 2026},
note = {3 July 2026}
}
NUWA 记录 · BibTeX
@misc{neo-glm52-report-nuwa,
title = {Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks — NUWA research record},
url = {https://whitzard.tech/nuwa/whitzard-index/work/neo-glm52-report/},
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--neo-glm52-report.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--neo-glm52-report.json.}
}
证据快照 2026-10-29.1 · 索引更新 · RSS
参考文献
- [1] Early Results on Evaluating GLM 5.2 for Behavioral and Persuasion Risks. 2026-07-03. 3 July 2026 ; Report masthead and stated evaluation methodology; §1 Behavioral; §2 Persuasion benchmarks