SevenTnewS

Microsoft Research

9 published articles

AI5 min read

CARE-X Medical AI

Mild aortic dilation: 12% caught on first reads, 93% with a measuring VLM

Aortic dilation is rarely quantified on chest X-rays, and mild cases get missed: 5 of 43 on initial reads. A tool-augmented VLM caught 40. New CARE-X research explains why radiology AI needs rulers, not just eyes.

2026-08-16

LLMs & Models3 min read

Reinforcement learning

Microsoft's experiential learning fix gives AI models a coach, not just a score

Experiential Learning repurposes the LLM-as-a-Judge into an LLM-as-a-Coach that extracts transferable knowledge from each response and internalizes it via on-policy context distillation, beating rubric-based RL on held-out tasks and reducing reward hacking.

2026-07-30

LLMs & ModelsFeatured4 min read

Neuroscience

The neuroscience AI that finally cracks open the brain's black box

Generative causal testing (GCT) distills LLM-based brain-prediction models into concise verbal explanations, then uses an LLM to write stories that causally test those claims in fMRI. In experiments, GCT confirmed known selectivity, teased apart neighboring place-processing regions, and discovered new prefrontal micro-regions tuned to concepts like dialogue and measurements.

2026-07-11

NLP & ML5 min read

Memory management

Bing found a 12,000-line allocator that breaks a 50-year-old trade-off

Microsoft Research's mimalloc allocator uses thousands of per-page free lists and a clever page-stealing technique to achieve both high concurrency and low memory overhead, breaking a trade-off that constrained systems for decades. A deep-dive from the RiSE group reveals the design.

2026-07-10

AI AgentsFeatured4 min read

AI Agents

Microsoft's bet on small models for agentic AI is about orchestration, not knowledge

Microsoft Research's MagenticLite project shows that small models can handle complex agentic tasks when orchestration, model design, and execution environment are codesigned. The release includes MagenticBrain, a 14B orchestrator, and Fara1.5, a computer-use model family that nearly doubles prior performance on web navigation benchmarks.

2026-07-09

Tools & Frameworks4 min read

Frameworks & Tools

Microsoft's Flint hides the chart boilerplate so AI agents stop drawing wrong axes

Microsoft Research introduced Flint, a visualization intermediate language that helps LLMs and AI agents create polished charts without hand-coding low-level parameters like scales and axis formatting. In a study across three models, Flint outperformed direct Vega-Lite generation, and is already used internally in Data Formulator.

2026-07-08

AI4 min read

Nature Neuroscience paper

Microsoft's new method turns black-box brain AI into readable theories

GCT translates uninterpretable LLM-based brain models into short phrases like 'food preparation' or 'location names,' then uses an LLM to write stories that causally test those explanations in real subjects. The method promises to bridge predictive AI and human-readable scientific theory.

2026-07-05

Open Source2 min read

Open source

Microsoft just opened a codebase that kills the worst part of enterprise data work

Data Formulator 0.7 is an open-source AI system from Microsoft Research that helps enterprise teams connect fragmented data sources, prepare and analyze data with context-aware agents, and iteratively refine visualizations in a shared workspace, all without requiring SQL or programming skills.

2026-07-03

AI4 min read

Memory Management

mimalloc, Microsoft's tiny memory workhorse, is quietly powering AI at scale

Microsoft Research's mimalloc memory allocator, designed for high concurrency and large memory scales, is quietly powering AI and cloud services at scale. Its thread-local design with thousands of free lists minimizes contention and delivers both speed and memory efficiency.

2026-07-03