Foundation Models Academic Research (AFMR) - circular background pattern

Accelerating Foundation Models Research

Engaging the broader community in reimagining computing research

ニュース&特集

background pattern
Microsoft Research ブログ

| Rianne van den Berg, Jan Hermann, Christopher Bishop, と Paola Gori Giorgi

Microsoft researchers achieved a breakthrough in the accuracy of DFT, a method for predicting the properties of molecules and materials, by using deep learning. This work can lead to better batteries, green fertilizers, precision drug discovery, and more.

The image shows a radar chart comparing the performance of different AI models across various metrics. The chart has a circular grid with labeled axes including VO, AS, CEc, CEe, CL, MCr, MCt, MCu, MS, QLI, QLqA, SNs, KNa, KNc, KNF, KNn, and AT. Different AI models are represented by various line styles: Babbage-002 (dotted line), Davinci-002 (dash-dotted line), GPT-3.5-Turbo (dashed line), GPT-4.0 (solid thin line), OpenAI ol-mini (solid thick line), and OpenAI o1 (solid bold line). There is a legend in the bottom left corner explaining the line styles for each model. The background transitions from blue on the left to green on the right.
Microsoft Research ブログ

| Lexin Zhou と Xing Xie

ADeLe, a new evaluation method, explains what AI systems are good at—and where they’re likely to fail. By breaking tasks into ability-based requirements, it has the potential to provide a clearer way to evaluate and predict AI model performance.

Microsoft Research Podcast | Ideas: Evelyne Viegas, Muhammed Idris, Cesar Torres
Microsoft Research ポッドキャスト

| Gretchen Huizinga, Evelyne Viegas, Muhammed Idris, と Cesar Torres Jr.

Innovative AI research often depends on access to resources. Microsoft wants to help. Technical Advisor Evelyne Viegas and distinguished faculty from two Minority Serving Institutions discuss the benefits of Microsoft’s Accelerating Foundation Models Research program in their lives and research.

Illustrated image of Jindong Wang and Steven Eujiong Whang. “Abstracts: A Microsoft Research Podcast” runs along the bottom.
Microsoft Research ポッドキャスト

| Gretchen Huizinga, Jindong Wang, と Steven Euijong Whang

Researcher Jindong Wang and Associate Professor Steven Euijong Whang explore the NeurIPS 2024 work ERBench. ERBench leverages relational databases to create LLM benchmarks that can verify model rationale via keywords in addition to checking answer correctness.

Illustrated images of Siddharth Suri and David Holtz. “Collaborators: A Microsoft Research Podcast” runs along the bottom.
Microsoft Research ポッドキャスト

| Gretchen Huizinga, David Holtz, と Siddharth Suri

Researcher Siddharth Suri and professor David Holtz give a brief history of prompt engineering, discuss the debate behind their recent collaboration, and share what they found from studying how people’s approaches to prompting change as models advance.

ニュース | Microsoft

Through our Accelerating Foundation Models Research program, we’ve made grants to hundreds of projects in AI safety and alignment research, AI-driven scientific discovery, and beneficial applications of AI. And we launched our Global Perspectives Responsible AI Fellowship program, designed to…

ニュース | Microsoft On the Issues

As a company at the forefront of AI research and technology, we are committed to sharing our practices with the public as they evolve. This report enables us to share our maturing practices, reflect on what we have learned, chart…

ニュース | Microsoft Source

In early 2023, Professor Alice Oh and her colleagues at the Korea Advanced Institute of Science and Technology (KAIST) realized they needed to address the quickly growing interest in OpenAI ChatGPT among KAIST’s students. ChatGPT – a tool developed by…

Microsoft Research ブログ

Microsoft announces the AFMR Minority Serving Institutions grant recipients, advancing AI research focused on today’s most significant technical and societal challenges. The grant provides funding and access to Azure-hosted foundation models.