RenderFormer: 神经网络如何重塑3D渲染？
编者按：在计算机图形学发展数十年之后，渲染技术正迎来一次由人工智能驱动的革新。传统的渲染方法依赖复杂的几何建模和物理光学公式，而神经网络渲染（neural rendering）则尝试以学习的方式建模光线传播过程。近期，微软亚洲研究院提出的 RenderFormer 模型首次展示了全新的图形渲染能力，可在不依赖传统光线追踪或光栅化的前提下，仅通过神经网络就能实现支持任意场景输入且具备全局光照效果的3...
How AI will accelerate biomedical research and discovery
| Peter Lee, Daphne Koller, Noubar Afeyan, and Dr. Eric Topol
Daphne Koller, Noubar Afeyan, and Dr. Eric Topol, leaders in AI-driven medicine, discuss how AI is changing biomedical research and discovery, from accelerating drug target identification and biotech R&D to helping pursue the “holy grail” of a virtual cell.
Phi-4-reasoning is a 14-billion parameter model specialized in complex reasoning tasks. It is trained using supervised finetuning (SFT) on diverse prompts and reasoning demonstrations from o3-mini. The model generates detailed reasoning chains and leverages inference-time compute effectively. Phi-4-reasoning-plus, an enhanced…
In the news | Matter
Closed-loop optimization using machine learning for the accelerated design of sustainable cements incorporating algal biomatter
The concrete industry is a large producer of global greenhouse gases, the majority of which are attributable to ordinary Portland cement. Our research demonstrates a sustainable alternative by incorporating whole macroalgae as a biomatter substitute in cement. Biomatter is a…
AI Testing and Evaluation: Learnings from pharmaceuticals and medical devices
| Kathleen Sullivan, Daniel Carpenter, Timo Minssen, and Chad Atalla
Professors Daniel Carpenter and Timo Minssen explore evolving pharma and medical device regulation, including the role of clinical trials, while Microsoft applied scientist Chad Atalla shares where AI governance stakeholders might find inspiration in the fields.
One of the driving forces behind AI’s rapid progress is access to large-scale, high-quality data, essential to enable training models to continuously improve and perform reliably. But that well is running dry. As the supply of usable internet data shrinks,…
UI-Evol: 让智能体“知行合一”
编者按：当 AI 智能体走出实验室，在真实世界中执行任务时，只有知识还远远不够。但如今的许多智能体即使拥有正确的信息，可在实际操作中仍频频失误，任务完成率令人堪忧。微软亚洲研究院的研究员对此提出了一个新方法 UI-Evol，旨在帮助计算机使用智能体（Computer Use Agent）进化知识，使智能体不仅“知道该怎么做”，还要真正“做得到”。在该方法的加持下，智能体在实际执行能力、稳定性和可靠...
In the news | AI Ireland
E204 ‘Multilingual Innovation and AI’ with Microsoft’s Kalika Bali
Today’s guest is Kalika Bali, Senior Principal Researcher at Microsoft Research India. In the episode, Kalika talks about her intriguing journey and impactful work, as well as peeling back the layers of how accidental moments can shape careers and how…
In the news | LinkedIn Article
Mapping global floods with 10 years of satellite radar data
Floods are among the deadliest and most costly forms of extreme weather, affecting millions of people and causing over $40 billion in damage annually. And they’re getting worse. As climate change increases the frequency and intensity of extreme precipitation, we…