一项题为“快速思考”的新研究论文探讨了前沿人工智能模型的内部推理能力，特别是它们在没有明确思维链（CoT）提示的情况下完成任务的能力。研究发现，在过去六年里，这些模型以50%的成功率完成任务的时间范围每年都在翻倍。例如，GPT-5.5现在可以在没有CoT的情况下花费超过3分钟完成任务，研究人员预测到2030年这一时间可能延长至25分钟，这引发了对人工智能安全监督的担忧。 AI
影响 这项研究突显了人工智能安全监控中一个潜在的盲点，因为模型越来越多地在没有明确步骤的情况下在内部执行复杂的推理。
排序理由 该集群包含一篇详细介绍人工智能模型能力研究结果的研究论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2606.07157v1 Announce Type: new Abstract: Many efforts to ensure frontier AI models are safe rely on monitoring their chain-of-thought (CoT) reasoning. If models become able to perform sufficiently complex reasoning internally, without explicit thinking tokens, this would u…
Many efforts to ensure frontier AI models are safe rely on monitoring their chain-of-thought (CoT) reasoning. If models become able to perform sufficiently complex reasoning internally, without explicit thinking tokens, this would undermine such oversight. We measure how well fro…