朝花夕拾

Improving Auto-Design of Neural PDE Solvers with a Domain-Specific Language

arXiv:2608.04384v1 Announce Type: new Abstract: Neural PDE solver auto-design is fundamentally a search-space representation problem. In the space of unrestricted Python programs, valid solvers form an extremely sparse s

NeuMoSync: End-to-End Neuromodulatory Control for Plasticity and Adaptability in Continual Learning

arXiv:2608.04358v1 Announce Type: new Abstract: Continual learning (CL) requires models to learn tasks sequentially, yet deep neural networks often suffer from plasticity loss and poor knowledge transfer, which can imped

The RAIL Principles for Neurosymbolic AI: Reasoning, Assurances, Interfacing and Learning

arXiv:2608.04285v1 Announce Type: new Abstract: Neurosymbolic AI systems that integrate machine learning and symbolic reasoning are rapidly gaining attention. They complement the data-intensive statistical approaches of

·AI/, ·arxiv.org ↗

MatrAIx: Simulating the World with 8.3 Billion Persona Agents

arXiv:2608.04205v1 Announce Type: new Abstract: Human evaluation of AI systems and digital products is costly, slow, and difficult to scale. Offline evaluations are more scalable but often abstract away human diversity a

·AI/, ·arxiv.org ↗

Adversarially Robust Abductive Fusion of Pre-trained Transformer-based Perception Models

arXiv:2608.04190v1 Announce Type: new Abstract: Deploying pre-trained perception models in novel environments degrades their accuracy under distributional shift, and assembling them alone does not recover it: combiners s

FinPerMA: A Theory-Informed, Event-Grounded Personalized-Memory Benchmark for LLM Agents

arXiv:2608.04095v1 Announce Type: new Abstract: Large language model (LLM) agents are increasingly used as personalized assistants in high-stakes domains such as financial advising, yet it remains unclear whether they ca

·AI/, , ·arxiv.org ↗

FinProBench: Evaluating Financial AI Agents with Role-Grounded Rubrics Derived from Professional Deliverables

arXiv:2608.04077v1 Announce Type: new Abstract: Evaluating financial AI agents requires criteria aligned with real professional work. Existing rubric methods typically derive criteria from task prompts or model outputs,

·AI/, ·arxiv.org ↗

The LLM Proposes, the Executive Disposes: A Self-Verifying Agent Instrument that Dissociates Commitment Drift from Binding Drift in Long-Horizon Agents

arXiv:2608.04066v1 Announce Type: new Abstract: How do you verify a long-horizon agent when its own state and self-reports are exactly what you cannot trust? We present an agent instrument built so that verification is s

·AI/, , ·arxiv.org ↗

A Long-Run Persistence Theory for AI Systems under the Redundancy-Adjusted Artificial Age Score

arXiv:2608.04012v1 Announce Type: new Abstract: Artificial intelligence systems are increasingly expected to operate over repeated cycles of interaction, adaptation, and update rather than through isolated one-shot outpu

阿里达摩院开放 15 项芯片与 AI 研究课题，启动“阿里星”顶尖人才招募

IT之家 8 月 6 日消息，阿里达摩院今日宣布了面向 2027 届毕业生的“阿里星”顶尖人才招募培养计划。达摩院今年已开放 15 项研究课题，涉及 AI 芯片、新型 CPU 架构、医疗多模态智能体、 AGI 决策等众多前沿探索。IT之家付具体课题如下：芯片方向 AI 芯片编程模型设计实现为自研 AI 芯片打造全新编程语言和编译器工具链，不是简单的 " 移植 LLVM 后端 "，而是覆盖 DSL 设计、编译器前端、中端优化框架

·AI/, , , ·ithome.com ↗

From asking to doing: How the world is putting ChatGPT to work

New OpenAI Signals data shows how people use ChatGPT worldwide, with country-level insights on adoption, usage trends, and evolving behavior.

·AI/·openai.com ↗

openai.com ↗

Wed, August 5 (7 stories)

别再吹AI生图了，不能图层编辑的AI都是“画饼”

兔展智能发布RabbitVis，让AI真正走完设计全流程

·AI/·qbitai.com ↗

qbitai.com ↗

消息称美国前沿人工智能模型网络安全审查机制暂不涉及开放权重模型

IT之家 8 月 5 日消息，综合《纽约时报》《华尔街日报》《Axios》、彭博社等外媒报道，美国政府在当地时间本周二与头部人工智能企业举行的闭门会议中表示，其今年 6 月宣布的前沿模型网络安全审查机制当前仅涉及闭源模型，开放权重模型至少暂时不会受到影响。知情人士透露，美国政府在会议中将需自愿接受审查的模型定义为具有最先进功能和国家安全风险的闭源模型，该机制的任何内容都不应被解释为限制已发布的开源模型。图源：Pexels

·AI/·ithome.com ↗

阿里云上线 One Key MCP 服务：兼容 Qoder、Codex 等，可一键调用多家 MCP 服务

IT之家 8 月 5 日消息，阿里云今日宣布上线 One Key MCP 服务。开发者可通过统一的阿里云百炼 API Key 调用所有生态伙伴 MCP 服务，兼容 Qoder、Codex、Claude Code、Cursor 等主流 Coding Agent ，简化多 MCP 服务的接入、鉴权与计费管理流程。此前开发者接入不同 MCP 服务需分别申请凭证、配置鉴权并完成联调；服务数量增加后，接入和维护成本也随之上升。 One Key

·AI/, , ·ithome.com ↗

卡帕西提出测试 AI 能力新思路：100 万 Tokens 额度构建《指环王》3D 交互场景

IT之家 8 月 5 日消息，OpenAI 联合创始人安德烈 · 卡帕西（Andrej Karpathy）于 8 月 2 日在 X 平台发布推文，提出一项 AI 基准测试思路：向模型提供《指环王》开篇首段文字，再要求模型创建对应的 3D 世界。在测试 AI 模型时，此前业内习惯使用生成“鹈鹕骑车图”的 SVG 图片方式，用来评估大语言模型在空间推理、几何逻辑和复杂代码生成方面的真实水平。不过卡帕西指出目前模型能力不断提升，等简单任

·AI/, ·ithome.com ↗

倒卖英伟达GPU算力比火箭卫星来钱快！马斯克交出SpaceX首份财报

二季度AI收入暴涨247%

·AI/, , ·qbitai.com ↗

qbitai.com ↗

Neurosymbolic Reasoning with Incremental Knowledge for Sample Efficient Hierarchical Reinforcement Learning

arXiv:2608.02993v1 Announce Type: new Abstract: (Flat) Reinforcement Learning (RL) agents face significant challenges in environments with sparse rewards that require long-horizon reasoning. A compelling approach to impr

·AI/, ·arxiv.org ↗

On the missing data layer and a potential solution

arXiv:2608.02949v1 Announce Type: new Abstract: Latin America is missing two foundational layers of AI infrastructure: the dataset layer and the benchmark layer. This paper targets the dataset layer. The dataset layer fa