（翻译）我们对 AI 智能的思考正确吗？
Computer scientist Melanie Mitchell discusses why artificial intelligence doesn’t “think” or “reason” like humans, and how we can create better methods for measuring machine cognition.
（翻译）为什么癌症药物对每个人的效果不同？
Why do cancer drugs work for some people but not others? New AI research looks at how tumor cells behave to better match treatments to patients.
（翻译）LLM智能体使用仿真模型进行受控实验
Abstract page for arXiv paper 2608.23622: LLM Agents Perform Controlled Experiments Using Simulation Models
（翻译）审计合成回忆录：对照其所描述生活的文档记录度量LLM生成自传中的场景级虚构
Abstract page for arXiv paper 2608.23640: Auditing the Synthetic Memoir: Measuring Scene-Level Confabulation in LLM-Generated Autobiography Against the Documented Record of the Life It Describes
（翻译）评价评价者：用于大语言模型评测的Rasch测量理论
Abstract page for arXiv paper 2608.27463: Rating the Raters: Rasch Measurement Theory for LLM Evaluation
（翻译）物联网与深度学习用于茶园中Postelectrotermes militaris检测与严重程度评估的有效性
arXiv:2608.27480v1 Announce Type: new Abstract: Tea plantations are vulnerable to Postelectrotermes militaris, commonly known as the Upcountry Live Wood Termite (ULWT), which can cause substantial damage when infestations remain undetected. This study proposes an IoT-enabled acoustic monitoring fram
（翻译）SCAFFOLD：面向计算机科学研究图表的大规模结构化数据集，包含图表问答与思维链推理轨迹
Abstract page for arXiv paper 2609.00018: SCAFFOLD: A Large-Scale Structured Dataset of Computer Science Research Figures with Diagram QA and Chain-of-Thought Reasoning Traces
（翻译）EvalDetectBench：衡量前沿语言模型评估意识的基准
Abstract page for arXiv paper 2609.01611: EvalDetectBench: A Benchmark for Measuring Evaluation Awareness in Frontier Language Models
Clair 想做“女性版连续血糖仪”，不扎针，用10类生物传感采集130多项指标，由机器学习反推激素与周期变化。文章拆解两位斯坦福毕业生如何用生理痕迹替代抽血，并规划 FDA 许可的两代产品。
（翻译）面向北极生态导航的自主 GeoAI 智能体
Abstract page for arXiv paper 2609.09374: An Autonomous GeoAI Agent for Arctic Eco-Navigation
当人工智能开始参与提出假设、设计实验，并根据实验结果调整研究方向，AI与科学研究的关系正在发生变化。从材料发现到生命系统模拟，再到创新药研发，AI for Science的探索正从单个环节的效率提升，走向对科研过程的系统性重构。
（翻译）用于事件预测的语言模型与先验的能力门控池化
Abstract page for arXiv paper 2609.12101: Competence-Gated Pooling of Language Models and Priors for Event Forecasting
