Multimodal models can drive unprecedented advances in AI, moving beyond text and numbers to a much wider spectrum of inputs. AI research leaders Peter Lee and Vijay Mital discussâ¯how a new generation of models can open new opportunities for businesses:⯠https://lnkd.in/ekufNZcJ
Microsoft Research
Think Tanks
Redmond, Washington 292,445 followers
We advance science and technology to benefit humanity.
About us
At Microsoft Research, we accelerate scientific discovery and technology innovation to empower every person and organization on the planet to achieve more. We do this by bringing together the best minds across diverse disciplines and backgrounds to take on the most pressing research challenges for Microsoft and for society. Our Research Lens We consider research directions through the lens of the positive impact we aspire to create with and for customers, communities, and all of society.
- Website
-
http://www.microsoft.com/research
External link for Microsoft Research
- Industry
- Think Tanks
- Company size
- 1,001-5,000 employees
- Headquarters
- Redmond, Washington
- Founded
- 1991
Updates
-
The healthcare industry has been an early mover in exploring emerging AI capabilities. AI research leaders Peter Lee and Vijay Mital discuss how Microsoft is helping, and how the lessons from healthcare could apply across many other industries. https://msft.it/6047mVS2X
-
How can we rigorously evaluate and understand state-of-the-art progress in AI? Eureka is an open-source framework for standardizing evaluations of large foundation models, beyond single-score reporting and rankings. Learn more about the extended findings. https://msft.it/6042mpZZQ
-
Investigating vulnerabilities in LLMs; A novel total-duration-aware (TDA) duration model for text-to-speech (TTS); Generative expert metric system through iterative prompt priming; Integrity protection in 5G fronthaul networks. https://msft.it/6040m3pW8
-
Medfuzz tests LLMs by breaking benchmark assumptions, exposing weaknesses vulnerabilities to bolster real-world accuracy. https://msft.it/6042mzAFC
-
Microsoft Research reposted this
Weâre advancing the Azure Quantum platform to tackle the worldâs most pressing challenges. We're announcing the best performing logical qubits on record with Quantinuum and will provide priority access to reliable quantum hardware in Azure Quantum with Atom Computing https://msft.it/6040mzafo. #AzureQuantum #Quantum
-
Peter Lee has been recognized by FierceHealthcare for his commitment to innovation, equity and improving lives. Read about his accomplishments at the Fierce 50 list. https://msft.it/6040mzzXQ
-
Microsoft Research reposted this
Measuring AI risk is both an art and a science, especially when adding context to the often fuzzy realm of human concepts. Explore how our expert team of researchers and technologists navigate complex ideas from linguistics to social sciences to create safe AI: https://msft.it/6003mG5lV
-
GraphRAG uses LLM-generated knowledge graphs to substantially improve complex Q&A over retrieval-augmented generation (RAG). Discover automatic tuning of GraphRAG for new datasets, making it more accurate and relevant. https://msft.it/6048mKmEM
-
The ML with New Compute Paradigms workshop at NeurIPS 2024âwhich will explore how emerging compute tech can sustainably scale amid demand for GPU-based AIâhas extended its call for papers through September 11. Learn more about the submission guidelines. https://lnkd.in/g2tgTE8q
The future of compute hardware for AI is more uncertain than ever. How should AI models change if compute hardware changes? What sort of models could be build with different, for example, analog or neuromorphic hardware? How should we train models that run on analog or photonic chips? How can models deal with noise or very narrow data types? We will discuss this and more at our NeurIPS 2024 workshop in Vancouver! Call for Papers: ML with New Compute Paradigms (MLNCP) Workshop at NeurIPS 2024 We call for workshop papers including but not limited to the following directions: ·       Advances in machine-learning that benefit from compute paradigms beyond standard digital compute, for example analog, photonic, in-memory, neuromorphic, or quantum compute. ·       Advances in machine learning methods that can handle challenges such as low bit precision, variability or noise induced by new hardware. Examples include inherently noise-robust models, algorithms to reduce errors in ML arithmetic, or ways to share and distribute models across unique analog hardware. ·       Advances in machine-learning paradigms that facilitate training and/or inference on new hardware paradigms. This can be general or specific to, for example, generative models for modalities such as image or sequence data. ·       Surveys and position papers for machine-learning with new compute paradigms. The submission deadline is Aug 29 '24 (Anywhere on Earth) Please check the workshop website for updates on format and possible deadline extensions. Workshop Website: www.mlwithnewcompute.com
ML with New Compute Paradigms (MLNCP) at NeurIPS 2023
mlwithnewcompute.com