rui 的 blog

菜单

Administrator
发布于 2026-05-19 / 3 阅读
0
0

AI 每日资讯 - 2026-05-19

发布日期：2026-05-19

收录条目：20

1. Meet MemPrivacy: An Edge-Cloud Framework that Uses Local Reversible Pseudonymization to Protect User Data Without Breaking Memory Utility

摘要：As LLM-powered agents move from research to production, one design tension is becoming harder to ignore: the more useful cloud-hosted memory becomes, the more private user data it exposes. Researchers from MemTensor (Sha

2. Stochastic Gradient Descent (SGD’s) Frequency Bias and How Adam Fixes It

摘要：Modern language models are trained on data with extremely uneven token distributions. A small number of words appear in almost every sentence, while many rare but meaningful tokens occur only occasionally. This creates a

3. Musk v. Altman proved that AI is led by the wrong people

摘要：The tech trial of the year, Musk v. Altman, was ultimately a fight for control. Elon Musk argued that Sam Altman, with whom he helped found the now-massive company OpenAI, shouldn't direct the future of AI. Altman's lawy

4. Prompting Amazon Nova 2 for content moderation

摘要：In this post, you learn how to prompt Amazon Nova 2 Lite for content moderation using structured and free-form approaches, grounded in the MLCommons AILuminate Assessment Standard. The prompting techniques use the AILumi

5. All of the updates from Elon Musk and Sam Altman’s battle over OpenAI

摘要：Sam Altman and Elon Musk are facing off in a high-stakes trial that could alter the future of OpenAI and its most well-known product, ChatGPT. In 2024, Musk filed a lawsuit accusing OpenAI of abandoning its founding miss

6. Elon Musk loses his case against Sam Altman

摘要：After around two hours of deliberation, the jury has reached a unanimous verdict in Musk v. Altman, the tech trial of the year. The group found that two claims were barred by the statute of limitations, and a third faile

7. Aderant transforms cloud operations with Amazon Quick

摘要：In this post, we share how Aderant used the AI-powered capabilities of Amazon Quick to unify search across six vendor systems and automate documentation workflows, achieving 90 percent faster search times and 75 percent

8. Amazon Alexa Plus can now create AI-generated podcasts

摘要：Alexa Plus, Amazon's upgraded AI assistant, can now generate podcasts on "virtually any topic," according to an announcement on Monday. With the update, Amazon says you can give Alexa Plus a topic, and the AI assistant w

9. Integrate Atlassian Confluence Cloud with Amazon Quick

摘要：In this post, you will learn how to set up the Confluence Cloud integration with Quick. This includes creating a knowledge base for semantic search, setting up Actions to query and manage Confluence pages, and organizing

10. Build custom code-based evaluators in Amazon Bedrock AgentCore

摘要：In this post, you will implement four Lambda-based custom code evaluators for a financial market-intelligence agent, register each with AgentCore, and run them in on-demand and online modes. You will also see how to comb

11. OpenAI and Dell partner to bring Codex to hybrid and on-premise enterprise environments

摘要：OpenAI and Dell partner to bring Codex to hybrid and on-premise environments, helping enterprises deploy AI coding agents securely across data and workflows.

12. NVIDIA Introduces a 4-Bit Pretraining Methodology Using NVFP4, Validated on a 12B Hybrid Mamba-Transformer at 10T Token Horizon

摘要：NVIDIA introduces a 4-bit pretraining methodology built around the NVFP4 microscaling format — combining selective BF16 layers, 16×16 Random Hadamard Transforms on Wgrad inputs, 2D weight scaling, and stochastic rounding

13. DeepSlide: From Artifacts to Presentation Delivery

摘要：arXiv:2605.15202v1 Announce Type: new Abstract: Presentations are a primary medium for scholarly communication, yet most AI slide generators optimize the artifact (a visually plausible deck) while under-optimizing the de

14. SDOF: Taming the Alignment Tax in Multi-Agent Orchestration with State-Constrained Dispatch

摘要：arXiv:2605.15204v1 Announce Type: new Abstract: Multi-agent orchestration frameworks such as LangChain, LangGraph, and CrewAI route tasks through graph-based pipelines but do not enforce the stage constraints that govern

15. Does Theory of Mind Improvement Really Benefit Human-AI Interactions? Empirical Findings from Interactive Evaluations

摘要：arXiv:2605.15205v1 Announce Type: new Abstract: Improving the Theory of Mind (ToM) capability of Large Language Models (LLMs) is crucial for effective social interactions between these AI models and humans. However, the

16. SkillSmith: Compiling Agent Skills into Boundary-Guided Runtime Interfaces

摘要：arXiv:2605.15215v1 Announce Type: new Abstract: Recently, skills have been widely adopted in large language model (LLM)-based agent systems across various domains. In existing frameworks, skills are typically injected in

17. Fair outputs, Biased Internals: Causal Potency and Asymmetry of Latent Bias in LLMs for High-Stakes Decisions

摘要：arXiv:2605.15217v1 Announce Type: new Abstract: Instruction-tuned language models exhibit behavioural fairness in high-stakes decisions while retaining biased associations in their internal representations. However, whet

18. CAX-Agent: A Lightweight Agent Harness for Reliable APDL Automation

摘要：arXiv:2605.15218v1 Announce Type: new Abstract: Large language models deployed for MAPDL finite-element simulation face practical reliability challenges: without structured execution control, tool encapsulation, and faul

19. NOVA: Fundamental Limits of Knowledge Discovery Through AI

摘要：arXiv:2605.15219v1 Announce Type: new Abstract: Can AI systems discover genuinely new knowledge through iterative self improvement, and if so, at what cost? We introduce the NOVA framework, which models the common ``gene

20. ICRL: Learning to Internalize Self-Critique with Reinforcement Learning

摘要：arXiv:2605.15224v1 Announce Type: new Abstract: Large language model-based agents make mistakes, yet critique can often guide the same model toward correct behavior. However, when critique is removed, the model may fail

NoSuchUpload 复盘：为...

评论