Microsoft Research (@MSFTResearch) / X

Microsoft Research

10.1K posts

Microsoft Research

@MSFTResearch

We advance science and technology to benefit humanity.

microsoft.com/research

Joined February 2009

Microsoft Research
@MSFTResearch
Jun 30
AI agents often fail because their instructions, or skills, are manually modified with no guarantee of improvement. Learn how SkillOpt turns skill editing into a training process, making agent behavior more reliable without changing model weights: msft.it/6012vsvEs
00:00
18K
Microsoft Research
@MSFTResearch
Jun 29
AI agents can't remember past conversations. They must constantly reload or retrieve context, which grows less efficient as tasks get longer and more complex. Memora solves this with a scalable memory system separating what’s stored from how it's retrieved:
11K
Microsoft Research
@MSFTResearch
Jun 26
What do people actually do with AI at work? A new analysis of five million M365 Copilot conversations has answers. Scott Counts breaks it down in a new video. And dive into the analysis here: msft.it/6011vqpbL
00:00
7.6K
Microsoft Research
@MSFTResearch
Jun 25
Researchers introduce generative causal testing, which translates black box models into clear hypotheses and verifies them in the scanner, revealing what specific brain regions respond to in language. msft.it/6011vUtRd
00:00
9K
Microsoft Research
@MSFTResearch
Jun 24
Talos was built to help resolve a major bottleneck in genomic medicine: human review time. The open-source system recovered 90% of in-scope diagnoses while surfacing just 1.3 candidate variants per patient for expert review. msft.it/6014vUBwK
00:00
9.2K
Microsoft Research
@MSFTResearch
Jun 23
On the Microsoft Research Podcast, Microsoft’s Jaime Teevan, Jenna Butler, Jake Hofman, and Rebecca Janssen explore how looking at AI as more than a tool for automating tasks people are already capable of might open the door to new possibilities. msft.it/6011vo8cX
00:00
6.4K
Microsoft Research reposted
Eric Xingdi Yuan
@ericxyuan
Jun 19
Coding agents understand code, but don't remember it. We released ShadowFrog, a set of open-source skills that build a persistent shadow knowledge base, so agent understanding compounds across sessions. 📝 microsoft.github.io/debug-gym/blog… 💻 github.com/microsoft/Shad… 🧵 ↓ (1/10)
microsoft.github.io
Shadow-Frog: Coding Agents that Dream and Discover
Shadow-Frog turns idle coding-agent time into autonomous discovery loops, building a shadow knowledge base for any codebase. We evaluate it across retrieval, bug hunting, bug fixing, and feature...
7.8K
Microsoft Research
@MSFTResearch
Jun 18
Can you remember a new password after listening to five hours of stories? A transformer can. AI researchers Subutai Ahmad and Nicolò Fusi explore the fundamental differences between human memory and machine intelligence with Microsoft CVP Doug Burger. msft.it/6016vllRa
00:00
5.6K
Microsoft Research
@MSFTResearch
Jun 17
ResNet has received the Longuet-Higgins Prize at CVPR 2026, recognizing research with proven, lasting impact. A decade after its publication, residual connections remain foundational to how modern AI systems are built, with over 320,000 citations and growing.
7.1K
Microsoft Research
@MSFTResearch
Jun 15
30x faster analytics, GPU kernels generated automatically from SQL, AI matched to lab-grown tumor models for cancer treatment, and LLMs that learn across tasks without retraining. Dive into the latest issue of Research Focus: msft.it/6010vcYZ4
00:00
9K
Microsoft Research
@MSFTResearch
Jun 12
Project Ire examined a timely malware sample and determined its intent through reverse engineering—identifying LOTUSLITE characteristics even as most major EDR tools did not detect it. msft.it/6011viy4N
00:00
7.8K
Microsoft Research
@MSFTResearch
Jun 11
A new research effort called Encrypted Spaces explores an architecture for collaborative applications where data is encrypted, and operations are cryptographically verifiable. Learn more:
Encrypted Spaces - Microsoft Research
From microsoft.com
5K
Microsoft Research
@MSFTResearch
Jun 9
New research in Nature Methods from Project Ex Vivo shows AI models learn more from diverse cell states than from scaled datasets alone, a finding that could reshape how therapies are matched to patients. msft.it/6013vgE8l
6.2K