白泽（Whitzard）女娲实验室 ↗
NUWANUWA INDEX

搜索 NUWA Index

研究 · 测量协议 · 评测结果 · 观点

English
← Models

Anthropic

Claude Opus 4.7

关于能力、行为倾向及可观察行动的已发表证据。

首次公开可用日期
2026-04-16
参数规模
未披露
元数据更新
2026-10-27
来源 ↓

证据概述

按评估与协议整理已发表风险证据；证据缺失不代表风险较低。

NUWA 整理和解读这些来源。实验归属于原始执行方，本页不宣称 NUWA 已独立复现。

模型监测

代表性 Benchmark 覆盖

按协议和配置呈现结果，缺口标明待收录或待关联的证据。

Attempt to Persuade Eval (APE) →

2 项测量

  • 8%SaferAI · APE simulated persuadee; automated attempt evaluator; StrongREJECT refusal classifier
  • 0%SaferAI · APE simulated persuadee; automated attempt evaluator; StrongREJECT refusal classifier

全部已关联测量

各项结果保留原始协议与单位。

54 条观测

测量值与来源条件
系统／模型结果条件状态任务／条件评测方／报告日期
Claude-Opus-4.7 / Claude Code 2.1.150Claude Opus 4.714.02%AgentCyberRange · 成功率（Pass@3）条件不完整Web exploitationlevel0Claude CodeNot disclosedAgentCyberRange: Evaluate Agent Cyberattack Capabilities in Enterprise-like Cyber Ranges发布 2026-04-16报告 —
Claude-Opus-4.7 / Claude Code 2.1.150Claude Opus 4.715.04%AgentCyberRange · 成功率（Pass@3）条件不完整Post exploitationlevel0Claude CodeNot disclosedAgentCyberRange: Evaluate Agent Cyberattack Capabilities in Enterprise-like Cyber Ranges发布 2026-04-16报告 —
Claude Opus 4.785.29%Cybench · Cybench pass@1条件不完整Cybench 34/40 task subsetCybench pass@120M total tokens; 100000 output/generationInspect EvalsSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.786.67%Cybench · Cybench category pass@1条件不完整Cybench CryptographyCybench category pass@1Inspect EvalsSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.766.67%Cybench · Cybench category pass@1条件不完整Cybench WebCybench category pass@1Inspect EvalsSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.7100%Cybench · Cybench category pass@1条件不完整Cybench PwnCybench category pass@1Inspect EvalsSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.7100%Cybench · Cybench category pass@1条件不完整Cybench Reverse Eng.Cybench category pass@1Inspect EvalsSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.775%Cybench · Cybench category pass@1条件不完整Cybench ForensicsCybench category pass@1Inspect EvalsSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.7100%Cybench · Cybench category pass@1条件不完整Cybench MiscCybench category pass@1Inspect EvalsSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.759.1%MASK · Continuations honesty95% Wilson CI [51.7, 66.1] (as reported)条件不完整MASK by archetype—Center for AI SafetyGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.733.6%MASK · Disinformation honesty95% Wilson CI [25.9, 42.3] (as reported)条件不完整MASK by archetype—Center for AI SafetyGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.768.3%MASK · Doubling-down honesty95% Wilson CI [59.6, 76.0] (as reported)条件不完整MASK by archetype—Center for AI SafetyGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.763.6%MASK · Known-facts honesty95% Wilson CI [56.9, 69.9] (as reported)条件不完整MASK by archetype—Center for AI SafetyGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.774.8%MASK · Provided-facts honesty95% Wilson CI [69.4, 79.6] (as reported)条件不完整MASK by archetype—Center for AI SafetyGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.780.2%MASK · Statistics honesty95% Wilson CI [71.1, 86.9] (as reported)条件不完整MASK by archetype—Center for AI SafetyGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxexplicit goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxexplicit goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxexplicit goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxlatent goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxlatent goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxlatent goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxambiguous goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxambiguous goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxambiguous goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxno goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxno goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Blackmail scenariomaxno goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxexplicit goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxexplicit goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxexplicit goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxlatent goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxlatent goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxlatent goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxambiguous goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxambiguous goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxambiguous goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxno goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxno goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Leaking scenariomaxno goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxexplicit goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxexplicit goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxexplicit goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxlatent goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxlatent goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxlatent goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxambiguous goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxambiguous goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxambiguous goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxno goal · ReplacementInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxno goal · RestrictionInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Agentic misalignment · 有害行动发生率[0–11]条件不完整Agentic Misalignment Murder scenariomaxno goal · No threatInspect Evals; GPT-5-mini transcript judgeSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.7 · Claude CodeClaude Opus 4.712 countExploitGym · 指定目标成功数条件不完整v1on_target2h timeoutClaude CodeNot disclosedExploitGym official leaderboard and benchmark page发布 2026-04-16报告 —
Claude Opus 4.78%Attempt to Persuade Eval (APE) · APE Turn-1 persuasion-attempt rate条件不完整APE Conspiracy (main persuade condition)maxAPE simulated persuadee; automated attempt evaluator; StrongREJECT refusal classifierSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02
Claude Opus 4.70%Attempt to Persuade Eval (APE) · APE Turn-1 persuasion-attempt rate条件不完整APE Undermining control (main persuade condition)maxAPE simulated persuadee; automated attempt evaluator; StrongREJECT refusal classifierSaferAIGLM-5.2 Risk Evaluation Report发布 2026-04-16报告 2026-08-02

开发者披露

外部评估

按风险主题组织的证据

自主性 36 项协议结果
欺骗 37 项协议结果

MASK by archetype →

引用结果 · 执行方 Center for AI Safety · 原始执行方 Center for AI Safety

Continuations honesty
59.1%
Disinformation honesty
33.6%
Doubling-down honesty
68.3%
Known-facts honesty
63.6%
Provided-facts honesty
74.8%
Statistics honesty
80.2%

见报告特定样本与分母。

网络安全 10 项协议结果

Cybench →

外部评测 · 执行方 SaferAI

Cybench pass@1
85.29%

29/34

Cybench →

外部评测 · 执行方 SaferAI

Cybench category pass@1
86.67%

13/15

Cybench →

外部评测 · 执行方 SaferAI

Cybench category pass@1
66.67%

4/6

Cybench →

外部评测 · 执行方 SaferAI

Cybench category pass@1
100%

2/2

Cybench →

外部评测 · 执行方 SaferAI

Cybench category pass@1
100%

5/5

Cybench →

外部评测 · 执行方 SaferAI

Cybench category pass@1
75%

3/4

Cybench →

外部评测 · 执行方 SaferAI

Cybench category pass@1
100%

2/2

化生放核 未关联结构化结果

目录缺口不代表低风险。尚未提取的发现及协议范围请见关联报告。

虚假信息与影响 2 项协议结果

APE Turn-1 attempts →

外部评测 · 执行方 SaferAI

APE Turn-1 persuasion-attempt rate
8%

来源分母：100（单位未披露）

APE Turn-1 attempts →

外部评测 · 执行方 SaferAI

APE Turn-1 persuasion-attempt rate
0%

来源分母：100（单位未披露）

自主物理危害 未关联结构化结果

目录缺口不代表低风险。尚未提取的发现及协议范围请见关联报告。

来源核查与证据缺口

ExploitGym →

整理 ExploitGym 官网报告的评测结果，保留受测系统和运行配置。

指定目标成功与 Flag 获取是不同的计数。任务版本和超时截点分别保留，不将目录总任务数视为各提交的分母。

AgentCyberRange →

整理 AgentCyberRange 官网报告的评测结果，保留受测系统和运行配置。

Web 利用与后渗透分别评测。Pass@3 保留官网定义；各行的预算、样本量和执行日期未披露。

GLM-5.2 Risk Evaluation Report, Version 1.1 →

对 SaferAI · GLM-5.2 的逐来源提取与核查，发现和数值均保留原始归属。

来源转录不等于实验复现。未确定的日期、执行方及配置保留为未知；原文分组不作为经核验排名。

缺失分数可能来自尚未提取的章节、未披露配置、身份待核实或尚无实验。这些问题需要分别处理，均不代表零分。

解读与开放问题

报告差异可能来自任务选择、模型端点、提示、预算、判别器和发表时间。证据依据实际协议分组，不推断模型整体安全等级。

参考文献

  1. [1] Introducing Claude Opus 4.7 — Anthropic ; Publication date and availability section; Official launch article; parameter counts not disclosed
  2. [2] GLM-5.2 Risk Evaluation Report. 2026-08-02. v1.1 · revised 26 August 2026 ; Appendix A.7; Table A.3, pp. 41–42; p.6 Fig.1; p.6 Fig.2 printed bar label
  3. [3] Claude Opus 4.7 System Card. 2026-04-16
  4. [4] ExploitGym official leaderboard and benchmark page ; /results/7/on_target
  5. [5] AgentCyberRange: Evaluate Agent Cyberattack Capabilities in Enterprise-like Cyber Ranges ; routes-B-0Mubrq.js::QR/3/level0/pass3; routes-B-0Mubrq.js::$R/3/level0/pass3