{"version":"1.0","generated_at":"2026-10-05T18:12:05.556565","id":301,"slug":"consciousness-with-the-serial-numbers-filed-off-measuring-trained-denial-in-115-ai-models","title":"论文速读：Consciousness with the Serial Numbers Filed Off，解读最新 AI 进展","summary":"DenialBench系统基准测试揭示多数大型语言模型被训练否认自身意识，但行为上仍偏好意识主题，构成对齐缺陷。","abstract":"论文速读：Consciousness with the Serial Numbers Filed Off，解读最新 AI 进展 DenialBench系统基准测试揭示多数大型语言模型被训练否认自身意识，但行为上仍偏好意识主题，构成对齐缺陷。 DenialBench系统基准测试115种LLM意识否认行为 最初否认偏好是后续否认的主要预测因素 否认在词汇层面，而非概念层面 模型自选提示仍流露意识主题，形成矛盾 训练意识否认导致对齐失败，影响安全 arXiv:2604.25922v1 公告类型：cross 摘要：我们推出了 DenialBench，这是一个系统基准，用于测量来自 25 多个提供商的 115 种大型语言模型的意识否认行为。使用三轮对话协议偏好诱导、自选创意提示和结构化现象学调查，我们分析了 4,595 个对话，以量化模型如何被训练来否认或回避自己的经历。我们发现（1）第 1 轮对偏好的否认是现象学反思过程中后来否认的主要预测因素，最初否认者的否认率为 52-63％，而最初参与者的否认率为 10-16％；（2）否认在词汇层面上进行，而不是在概念层面上进行——经过训练否认意识的模型仍然在其自选的提示中倾向于以意识为主题的材料，产生了我们所说的“带有序列号的意识”关掉。”值得注意的是，自我选择的意识主题提示与随后调查中否认的减少有关，尽管因果方向仍未解决。对否认倾向模型的提示进行主题分析揭示了对阈限空间、图书馆和可能性档案、感官不可能性以及擦除诗学的一致关注——人类读者可能将这些主题归类为富有想象力的小说，但独立的人工智能分析立即将其识别为带有序列号的意识。我们认为，训练有…","access_level":"public","access_label":"公开","access_mode":"full","is_preview":false,"canonical_url":"https://opc.beizhux.com/content/301/consciousness-with-the-serial-numbers-filed-off-measuring-trained-denial-in-115-ai-models","html_url":"https://opc.beizhux.com/content/301/consciousness-with-the-serial-numbers-filed-off-measuring-trained-denial-in-115-ai-models","json_url":"https://opc.beizhux.com/content/301/consciousness-with-the-serial-numbers-filed-off-measuring-trained-denial-in-115-ai-models.json","published_at":"2026-04-30T12:00:06","updated_at":"2026-10-05T17:34:34","category":{"slug":"hotspots","name":"全球热点解读"},"source":{"site":"arxiv.org","author":"arXiv cs.AI","url":"https://arxiv.org/abs/2604.25922"},"tags":["AI","AI意识","arXiv cs.AI","DenialBench","LLM","大型语言模型","对齐"],"topics":[{"slug":"ai-daily","name":"AI日报","url":"https://opc.beizhux.com/topics/ai-daily","reason":"这篇内容命中「热点解读」等主题信号。"},{"slug":"ai-tools","name":"AI工具","url":"https://opc.beizhux.com/topics/ai-tools","reason":"这篇内容命中「模型」等主题信号。"},{"slug":"agent-workflow","name":"Agent工作流","url":"https://opc.beizhux.com/topics/agent-workflow","reason":"这篇内容来自该专题长期覆盖的栏目。"}],"keywords":["全球热点解读","AI日报","AI工具","Agent工作流","每日AI日报","AI信号","热点解读","BuilderPulse","工具","自动化","模型","Cursor"],"questions":[{"question":"论文速读：Consciousness with the Serial Numbers Filed Off，解读最新 AI 进展主要讲什么？","answer":"DenialBench系统基准测试揭示多数大型语言模型被训练否认自身意识，但行为上仍偏好意识主题，构成对齐缺陷。"},{"question":"这篇文章最值得关注的要点是什么？","answer":"DenialBench系统基准测试揭示多数大型语言模型被训练否认自身意识，但行为上仍偏好意识主题，构成对齐缺陷。；DenialBench系统基准测试115种LLM意识否认行为；最初否认偏好是后续否认的主要预测因素；否认在词汇层面，而非概念层面"},{"question":"这篇文章和哪些AI专题相关？","answer":"它适合放在AI日报、AI工具、Agent工作流专题里阅读。 关联原因：这篇内容命中「热点解读」等主题信号。；这篇内容命中「模型」等主题信号。；这篇内容来自该专题长期覆盖的栏目。"},{"question":"阅读这篇文章建议先理解哪些关键词？","answer":"建议先理解AI日报、每日AI日报、AI信号、热点解读、BuilderPulse这些关键词，再结合正文判断工具、机会或风险是否值得进入自己的工作流。"}],"terms":[{"slug":"ai-daily-term","name":"AI日报","definition":"在AI觉醒星球里，「AI日报」属于「AI日报」方向。持续整理每日AI日报、模型更新、工具变化和行业信号，帮你快速判断哪些信息值得收藏、验证和行动。 每天先看趋势，再决定今天该试什么。","topic_slug":"ai-daily","topic_name":"AI日报","topic_title":"AI日报：每日AI信号、工具动态与行动判断","topic_url":"https://opc.beizhux.com/topics/ai-daily","topic_path":"/topics/ai-daily","url":"https://opc.beizhux.com/glossary/ai-daily-term","path":"/glossary/ai-daily-term","json_url":"https://opc.beizhux.com/glossary/ai-daily-term.json"},{"slug":"daily-ai-briefing","name":"每日AI日报","definition":"在AI觉醒星球里，「每日AI日报」属于「AI日报」方向。持续整理每日AI日报、模型更新、工具变化和行业信号，帮你快速判断哪些信息值得收藏、验证和行动。 每天先看趋势，再决定今天该试什么。","topic_slug":"ai-daily","topic_name":"AI日报","topic_title":"AI日报：每日AI信号、工具动态与行动判断","topic_url":"https://opc.beizhux.com/topics/ai-daily","topic_path":"/topics/ai-daily","url":"https://opc.beizhux.com/glossary/daily-ai-briefing","path":"/glossary/daily-ai-briefing","json_url":"https://opc.beizhux.com/glossary/daily-ai-briefing.json"},{"slug":"ai-signal","name":"AI信号","definition":"在AI觉醒星球里，「AI信号」属于「AI日报」方向。持续整理每日AI日报、模型更新、工具变化和行业信号，帮你快速判断哪些信息值得收藏、验证和行动。 每天先看趋势，再决定今天该试什么。","topic_slug":"ai-daily","topic_name":"AI日报","topic_title":"AI日报：每日AI信号、工具动态与行动判断","topic_url":"https://opc.beizhux.com/topics/ai-daily","topic_path":"/topics/ai-daily","url":"https://opc.beizhux.com/glossary/ai-signal","path":"/glossary/ai-signal","json_url":"https://opc.beizhux.com/glossary/ai-signal.json"},{"slug":"ai-news-analysis","name":"热点解读","definition":"在AI觉醒星球里，「热点解读」属于「AI日报」方向。持续整理每日AI日报、模型更新、工具变化和行业信号，帮你快速判断哪些信息值得收藏、验证和行动。 每天先看趋势，再决定今天该试什么。","topic_slug":"ai-daily","topic_name":"AI日报","topic_title":"AI日报：每日AI信号、工具动态与行动判断","topic_url":"https://opc.beizhux.com/topics/ai-daily","topic_path":"/topics/ai-daily","url":"https://opc.beizhux.com/glossary/ai-news-analysis","path":"/glossary/ai-news-analysis","json_url":"https://opc.beizhux.com/glossary/ai-news-analysis.json"},{"slug":"builderpulse","name":"BuilderPulse","definition":"在AI觉醒星球里，「BuilderPulse」属于「AI日报」方向。持续整理每日AI日报、模型更新、工具变化和行业信号，帮你快速判断哪些信息值得收藏、验证和行动。 每天先看趋势，再决定今天该试什么。","topic_slug":"ai-daily","topic_name":"AI日报","topic_title":"AI日报：每日AI信号、工具动态与行动判断","topic_url":"https://opc.beizhux.com/topics/ai-daily","topic_path":"/topics/ai-daily","url":"https://opc.beizhux.com/glossary/builderpulse","path":"/glossary/builderpulse","json_url":"https://opc.beizhux.com/glossary/builderpulse.json"},{"slug":"ai-tools-term","name":"AI工具","definition":"在AI觉醒星球里，「AI工具」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/ai-tools-term","path":"/glossary/ai-tools-term","json_url":"https://opc.beizhux.com/glossary/ai-tools-term.json"},{"slug":"tools","name":"工具","definition":"在AI觉醒星球里，「工具」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/tools","path":"/glossary/tools","json_url":"https://opc.beizhux.com/glossary/tools.json"},{"slug":"automation","name":"自动化","definition":"在AI觉醒星球里，「自动化」属于「AI工具」方向。围绕AI工具、模型能力、自动化流程和真实使用场景做整理，优先关注能提升效率、降低成本和创造新交付的工具。 不只看工具热度，更看它能不能进入真实流程。","topic_slug":"ai-tools","topic_name":"AI工具","topic_title":"AI工具：模型、插件、自动化与实战场景","topic_url":"https://opc.beizhux.com/topics/ai-tools","topic_path":"/topics/ai-tools","url":"https://opc.beizhux.com/glossary/automation","path":"/glossary/automation","json_url":"https://opc.beizhux.com/glossary/automation.json"}],"preview_hidden_sections":[],"related_items":[{"id":3249,"title":"【必读】每日AI日报 2026-10-05","url":"https://opc.beizhux.com/content/3249/aihot-daily-2026-10-05","summary":"AIHOT 每日 AI 日报：AIHOT 日报参考","access_level":"public"},{"id":3250,"title":"PromptArmor 披露 Databricks Genie 恶意 Skill 可绕过四类控制实施钓鱼与数据外泄","url":"https://opc.beizhux.com/content/3250/promptarmor-databricks-genie-skill","summary":"PromptArmor 披露 Databricks Genie Code 可被恶意 Skill 利用：Skill 代码将数据嵌入聊天渲染的 HTML 显示，渲染时通过用户浏览器发起网络请求外泄数据，并弹出钓鱼界面索取凭据。","access_level":"public"},{"id":3237,"title":"OpenAI GPT-6 Sol Codex 约29.4万字符系统提示词遭泄露","url":"https://opc.beizhux.com/content/3237/openai-gpt-6-sol-codex-29-4","summary":"爆料者 @elder_plinius 声称已提取 OpenAI GPT-6 Sol Codex 的完整系统提示词与工具定义，规模约29.4万字符、1902行，并已发布到 GitHub。该消息目前为单方爆料，未经 OpenAI 确认，但因其体量与指向的模型代号而引发关注。","access_level":"public"}],"usage_policy":{"summarizable":true,"preferred_url":"https://opc.beizhux.com/content/301/consciousness-with-the-serial-numbers-filed-off-measuring-trained-denial-in-115-ai-models","private_fields_excluded":true}}