编者按：近日，微软研究院上线了面向全球研究界的全新线上系列活动 Microsoft Research Forum，旨在共同探讨人工智能时代的最新研究进展、大胆新颖的想法以及全球研究界关注的重要议题。来自微软研究院全球各地的研究人员将分享他们的研究洞见，并与大家进行在线讨论，希望碰撞出更多新的思想火花。 在最新一期的 Microsoft Research Forum 中，来自微软研究院不同实验室的研...
Ideas: Solving network management puzzles with Behnaz Arzani
| Gretchen Huizinga and Behnaz Arzani
Behnaz Arzani loves hard problems and the freedom to explore. That makes research a great fit! She discusses her work in network management, including the potential role of LLMs in the field; the challenges that excite her; and how storytelling…
In the news | The JoongAng
Microsoft’s secret weapon – research leader Peter Lee
Peter Lee, president of Microsoft Research, is a leading force in Microsoft's leap forward in the era of generative AI.
Research Focus: Week of June 10, 2024
In this issue: RELEVANCE automatically evaluates creative LLM responses; Recyclable vitrimer-based printed circuit boards; Lean Attention: Hardware-aware scalable attention mechanism; WaveCoder: a fine-tuned code LLM; New AutoGen training course.
SIBYL: A machine learning-based framework for forecasting dynamic workloads
| Rana Alotaibi, Hanxian Huang, Tarique Siddiqui, Carlo Curino, Jesús Camacho Rodríguez, and Yuanyuan Tian
SIBYL is a machine learning model that makes highly accurate predictions of database queries, enabling tuning for more efficiency. Applying traditional database optimizations to these predicted queries helps maintain high performance as demands change.
In the news | BBC Technology of Business
‘Insane’ amounts of data spurs new storage tech
Project Silica uses powerful lasers to enable a piece of glass about the size of a DVD to store more than seven terabytes of data, helping to manage the rapidly growing supply.
LST-Bench: A new benchmark tool for open table formats in the data lake
| Jesús Camacho Rodríguez, Ashvin Agrawal, Anja Gruenheid, Ashit Gosalia, Cristian Petculescu, Josep Aguilar-Saborit, Avrilia Floratou, Carlo Curino, and Raghu Ramakrishnan
LST-Bench is a new open-source benchmark designed to evaluate table formats in cloud environments. It extends existing benchmarks to better reflect real-world usage & performance of data lakes and easily integrates with commonly used analytical engines.
对话Nature子刊论文作者：DiG如何揭示蛋白质秘密
作者：科学智能中心 编者按：尽管当前利用人工智能技术预测生物分子结构的模型已经可以精确预测包括蛋白质、核酸、小分子、离子和修饰残基在内的复合物结构，但对于科学家们来说仅了解分子的微观结构还远远不够，因为分子的宏观属性和功能往往取决于分子结构在平衡态下的分布。 用于分子结构平衡分布预测的深度学习框架 Distributional Graphormer（DiG）的最新论文，近期在《自然-机器智能》（N...
重塑认知科学中的“行为”理解，提升智能体决策能力
作者：韩东起 编者按：随着科技的不断进步，人工智能与认知科学的交叉研究已成为推动科学前沿的重要力量。其中，习惯性行为与目标导向行为的相互作用是心理学、认知神经科学和人工智能等领域都十分关注的研究课题，因为从简单的日常习惯到复杂的决策制定过程，都是这两种行为模式交互的体现。为了更好地理解二者间的关系，微软亚洲研究院的研究员们通过引入贝叶斯“意图”变量，将习惯性行为与目标导向行为进行了整合，达到了计算...