1
Model Collapse as Cultural Evolution
从文化演化理论解释大模型自我训练导致的模型崩溃，提出五个可证伪预测，填补语言学空白。
arXiv:2605.23054v1 Announce Type: cross Abstract: Model collapse, the progressive degradation of LLMs trained on their own outputs, has been character…
2
摩根大通：预计标普500指数未来一年内将突破9000点，AI超级周期或比想象更牛
摩根大通大胆预测标普500突破9000点，背后AI超级周期才是真正的超级引擎。
摩根大通分析师在最新报告重表示，预计标普500指数在未来一年内将突破9000点，这意味着较当前水平上涨22%。摩根大通全球投资策略师克里蒂·古普塔（Kriti Gupta）等人在最新报告中指出，尽管目前市场上，仍存在对美国通货膨胀和伊朗战争的普遍担忧，但若人工智能超级周期的规模超出市场此前的预期，那…
3
LLM Retrieval for Stable and Predictable Ad Recommendations
LLM检索让广告推荐更稳定可预测，看论文如何用大模型优化推荐系统
arXiv:2605.21969v1 Announce Type: cross Abstract: Traditional ads recommendation systems have primarily focused on optimizing for prediction accuracy …
4
Conformal Selective Acting: Anytime-Valid Risk Control for RLVR-Trained LLMs
新方法保障LLM在线部署每轮风险可控，基于共形预测与RLVR训练，安全认证更可靠。
arXiv:2605.20270v1 Announce Type: new Abstract: A local specialist LLM, fine-tuned with reinforcement learning from verifiable rewards (RLVR) on opera…
5
AI-Augmented Surveys: Leveraging Large Language Models and Surveys for Opinion Prediction
用大语言模型预测民意调查缺失数据，弥补传统调查的历史局限，开辟AI+社会科学新路径。
arXiv:2305.09620v4 Announce Type: replace-cross Abstract: Nationally representative surveys track public opinion, yet they ask only a limited set of q…
6
Anthropic 有望 2026Q2 实现首个盈利季度，跑赢 OpenAI 与 xAI
Anthropic营收翻倍，预计2026Q2实现盈利，商业表现超越OpenAI与xAI。
IT之家 5 月 21 日消息，金融时报今天（5 月 21 日）发布博文，报道称 Anthropic 有望在 2026 年第 2 季度实现首个盈利季度 ，营收达 109 亿美元（IT之家注：现汇率约合 742.5 亿元人民币），较 Q1 的 48 亿美元翻倍增长，营业利润约 5.59 亿美元（现汇率…
7
Surviving the Unseen: Predictive Defense for Novel Multi-Turn Multimodal Attacks
多轮多模态攻击防不胜防？这篇论文提出预测性防御机制，应对未见过的新型攻击。
arXiv:2605.18988v1 Announce Type: cross Abstract: The expansion of Multimodal Large Language Models (MLLMs) and their integration into autonomous agen…
8
机构预测：谷歌 Android XR 智能眼镜今年销量将达 200 万台
谷歌Android XR智能眼镜走时尚路线，机构预测2026年销量达200万台，三星联名款即将亮相。
IT之家 5 月 21 日消息，谷歌首款智能眼镜产品遭遇了彻底失败，为避免重蹈覆辙，谷歌如今优先主打时尚设计路线，由眼镜品牌负责外观设计，三星承担内部硬件研发。谷歌则把控整机使用体验，重点依托 Gemini 大模型。目前尚难判断这套合作模式能否为各方带来收益，但早期市场预测认为其具备成功可能性。 谷…
9
The AI people have been right a lot
盘点那些年AI专家们的神准预言，看看他们如何一次次押中技术走向。
Article URL: https://dylanmatthews.substack.com/p/the-ai-people-have-been-right-a-lot Comments URL: https://news.ycombinator.com/item?id=48211361 Poin…
10
Counterpoint：2026 年全球 OLED 显示设备总支出预计同比增长 77%
2026年全球OLED设备支出预计激增77%，LCD投资微增，显示产业格局加速转向OLED。
IT之家 5 月 21 日消息，Counterpoint Research 在其当地时间 19 日发布的见解中预测， 2026 年全球 OLED 显示设备总支出将增长 77% 。在计算 LCD 设备投资微幅增长 (+1%) 的稀释后，全显示产业设备总支出同比增幅为 53%。 机构认为在 2023~2…
11
Toward an AI-Powered Computational Testbed for Workforce Policy
用AI构建劳动力政策模拟平台，提前预测人机协作的转型挑战。
arXiv:2605.19064v1 Announce Type: cross Abstract: Workforce transformations are difficult to forecast and costly to mismanage. In particular, the inte…
12
Self-Supervised Bootstrapping of Action-Predictive Embodied Reasoning
RSS 2026 论文提出自监督方法，让机器人通过预测行动自主学习推理，无需大量标注数据。
arXiv:2602.08167v2 Announce Type: replace-cross Abstract: Embodied Chain-of-Thought (CoT) reasoning has significantly enhanced Vision-Language-Action …
13
PhysioSeq2Seq: A Hybrid Physiological Digital Twin and Sequence-to-Sequence LSTM for Long-Horizon Glucose Forecasting in Type 1 Diabetes
用混合数字孪生+LSTM实现1型糖尿病长期血糖预测，精准度突破现有方法。
arXiv:2605.16860v1 Announce Type: new Abstract: Accurate long-horizon glucose forecasting is critical for automated insulin delivery systems, which he…
14
Predictable Confabulations: Factual Recall by LLMs Scales with Model Size and Topic Frequency
研究发现LLM的幻觉并非随机，而是与模型规模和主题频率呈可预测的比例关系。
arXiv:2605.18732v1 Announce Type: cross Abstract: While scaling laws govern aggregate large language model performance, no scaling law has linked fact…
15
LangPrecip: Language-Aware Multimodal Precipitation Nowcasting
用语言感知提升降水短临预报精度，结合多模态信息解决运动不明确难题
arXiv:2512.22317v3 Announce Type: replace-cross Abstract: Short-term precipitation nowcasting is an inherently uncertain and under-constrained spatiot…
16
Reconciling Enterprise AI Revenue
不同权威机构对2025年全球AI市场规模的预测从数千亿到万亿不等，为何差距如此巨大？一文拆解数字背后的定义边界与统计逻辑。
Article URL: https://philippdubach.com/posts/reconciling-enterprise-ai-revenue/ Comments URL: https://news.ycombinator.com/item?id=48208134 Points: 1 …
17
Bayesian Joint Model of Multi-Sensor and Failure Event Data for Multi-Mode Failure Prediction
提出贝叶斯联合模型融合多传感器与故障事件数据，实现对多模式故障的精准预测。
arXiv:2506.17036v2 Announce Type: replace-cross Abstract: Modern industrial systems are often subject to multiple failure modes, and their conditions …
18
From Division to Decision: Leveraging Temporal Cell-Stage Segmentation for Embryo Transferability Prediction
利用时序细胞阶段分割技术，精准预测胚胎移植成功率，为辅助生殖提供AI新方案。
arXiv:2605.18923v1 Announce Type: cross Abstract: Accurate selection of bovine embryos is a challenging task, as current practice relies on a single e…
19
PASC: Pipeline-Aware Conformal Prediction with Joint Coverage Guarantees for Multi-Stage NLP and LLM Pipelines
针对NLP和LLM多阶段流水线，提出管道感知的共形预测方法，实现联合覆盖保证，提升推理可靠性。
arXiv:2605.18812v1 Announce Type: new Abstract: Modern NLP and LLM systems are pipelines: named entity recognition (NER) -> entity disambiguation (NED…
20
Leveraging Graph Structure in Seq2Seq Models for Knowledge Graph Link Prediction
融合图结构与Seq2Seq模型，提出GA-S2S框架提升知识图谱链接预测效果，论文亮点突出。
arXiv:2605.18211v1 Announce Type: new Abstract: We introduce Graph-Augmented Sequence-to-Sequence (GA-S2S), a novel framework that integrates a T5-sma…