AI资讯日报 - 2026/7/13

👨‍🔬 Yiming Zhang, Zhonghan Zhao, Wenwei Zhang, Haiteng Zhao, Tianyang Lin, Yunhua Zhou, Demin Song, Kuikun Liu, Haochen Ye, Haian Huang, Yuzhe Gu, Haijun Lv, Qipeng Guo, Bin Liu, Gaoang Wang, Kai Chen

Evolution of Accuracy and Visual-Cognitive Errors in a Decade of Vision-Language AI Models

学术论文 ArXiv 重要度: 8

评估十年视觉语言模型进展，多模态大模型消除与人类描述差距，但空间依赖错误仍存。

👨‍🔬 Shravan Murlidaran, Miguel P. Eckstein

VEXAIoT: Autonomous IoT Vulnerability EXploitation using AI Agents

学术论文 ArXiv 重要度: 8

提出LLM多智能体框架自动挖掘IoT漏洞，攻击成功率高达95%，平均执行时间低于2分钟。

👨‍🔬 Katherine Swinea, Kshitiz Aryal, Lopamudra Praharaj, Maanak Gupta

ConceptSMILE: Auditing the Trustworthiness of Concept-Based Explainable AI

学术论文 ArXiv 重要度: 7

提出扰动审计框架评估概念可解释AI的可信度，发现不同概念路径可靠性差异。

👨‍🔬 Mohadeseh Mollapour, Koorosh Aslansefat, Zeinab Dehghani, Bhupesh Kumar Mishra, Tejal Shah, Zhibao Mian

Semantic Pareto-DQN: A Multi-Objective Reinforcement Learning Framework for Financial Anomaly Detection

学术论文 ArXiv 重要度: 7

多目标强化学习框架结合LLM语义表示，破解金融异常检测中的零召回陷阱。

👨‍🔬 Cláudio Lúcio do Val Lopes, Lucca Machado da Silva

Agora: Enhancing LLM Agent Reasoning Via Auction-Based Task Allocation

学术论文 ArXiv 重要度: 7

引入拍卖机制动态分配任务给专家模型，提升LLM代理推理效率与成本控制。

👨‍🔬 Kaiji Zhou, Ales Leonardis, Yue Feng

Lean-QIT: Towards a Formal Infrastructure for Quantum Information Theory

学术论文 ArXiv 重要度: 6

用Lean 4形式化量子信息论，验证量子编码定理，为AI辅助形式化提供基础。

👨‍🔬 Chengkai Zhu, Ziao Tang, Guocheng Zhen, Yimeng Cao, Yusheng Zhao, Ranyiliu Chen, Xuanqiang Zhao, Lei Zhang, Xin Wang

4DR360: State Reasoning for Joint 3D Detection and Occupancy Prediction in 4D Radar-Camera Full-Scene Perception

学术论文 ArXiv 重要度: 6

提出4D雷达-相机联合检测与占用预测框架，通过状态推理提升全场景感知能力。

👨‍🔬 Xiaokai Bai, Lianqing Zheng, Runwei Guan, Songkai Wang, Siyuan Cao, Hui-liang Shen

Task-Specific Multimodal Question Answering Agents via Confidence Calibration and Incremental Reasoning for QANTA 2026

学术论文 ArXiv 重要度: 6

针对多模态问答任务设计双智能体架构，通过置信度校准和增量推理在QANTA 2026中夺冠。

👨‍🔬 Nirjhar Das, Md. Al-Mamun Provath

PAC-ACT: Post-training Actor-Critic for Action Chunking Transformers

学术论文 ArXiv 重要度: 6

后训练强化学习框架优化动作分块变换器，提升工业接触式任务成功率并降低峰值接触力。

👨‍🔬 Yujie Pang, Zudong Li

TrustX Agent Risk Classification Framework (ARC): Risk-Tiering Internally Created Agentic AI Systems

学术论文 ArXiv 重要度: 5

提出智能体AI风险分类框架，含12维评分体系，用于内部代理系统的风险分层与治理。

👨‍🔬 Hannah M. Liu, Rhea Saxena, Shiv Asthana

🤖 AI资讯日报

📊 今日趋势总结

MIT Non-AI License

Why Boring Businesses Outlast AI Hype Cycles

Ask HN: Is the rate of progress in AI exponential?

Ask HN: Anyone concerned about NYC Local Law 144?

Ask HN: What's the pain using current AI algorithms?

The AI Crackpot Index

Ask HN: What would you read to learn about 'artificial intelligence'?

Common Lisp + Machine Learning Internship at Google

NLP, AI, ML, bots – a passing trend or much more?

Show HN: Startup Raising capital through Book Sales

The Next Bill Gates or Albert Einstein in AI 'Chris Clark'

Bioinformatician

PHINN-EEG: Topological Time-Series Analysis of Dream-State EEG -- Dynamic Betti Curves for Dream Content Classification and Topology-Conditioned Neural Signal Synthesis

Scalable Visual Pretraining for Language Intelligence

Evolution of Accuracy and Visual-Cognitive Errors in a Decade of Vision-Language AI Models

VEXAIoT: Autonomous IoT Vulnerability EXploitation using AI Agents

ConceptSMILE: Auditing the Trustworthiness of Concept-Based Explainable AI

Semantic Pareto-DQN: A Multi-Objective Reinforcement Learning Framework for Financial Anomaly Detection

Agora: Enhancing LLM Agent Reasoning Via Auction-Based Task Allocation

Lean-QIT: Towards a Formal Infrastructure for Quantum Information Theory

4DR360: State Reasoning for Joint 3D Detection and Occupancy Prediction in 4D Radar-Camera Full-Scene Perception

Task-Specific Multimodal Question Answering Agents via Confidence Calibration and Incremental Reasoning for QANTA 2026

PAC-ACT: Post-training Actor-Critic for Action Chunking Transformers

TrustX Agent Risk Classification Framework (ARC): Risk-Tiering Internally Created Agentic AI Systems

📅 历史日报目录