MLCommons

On Tuesday, October 15th, tune in to EDM Council's "Mitigating Generative AI Risks" panel. Discover effective strategies to address challenges such as hallucinations, unpredictable outputs, and ethical concerns in AI deployment. Don’t miss this opportunity to enhance your understanding of responsible AI practices! 

Join Us at MICCAI 2024! We’re excited to invite you to our tutorial on the Comprehensive Open Federated Ecosystem (COFE) at MICCAI on October 10th at MICCAI!   Discover the latest advancements in federated learning and how COFE is revolutionizing the field. Don’t miss this opportunity to learn, network, and innovate!   More details: https://lnkd.in/g3JmdJE2 #MICCAI2024 #FederatedLearning #COFE #AI #MachineLearning Spyridon Bakas, Qi Dou, Sarthak Pati, Siddhesh Thakur, Cyril Zakka, MD, Patrick Foley, Hasan Kassem IU Division of Computational Pathology, The Chinese University of Hong Kong, Hugging Face, Intel AI

Read the latest from HPCwire on MLCommons recent Storage benchmark. "Overall, the MLPerf Storage v1.0 benchmark is an excellent addition to the entire MLComons suite of tests. " https://lnkd.in/gFqyws7K

Congratulations Pooria Namyar on being named an MLCommons Rising Star. https://lnkd.in/gJ75_tYJ

MLCommons we believe that AI safety must be grounded in scientific understanding, and it's great to see so many embracing this approach, thanks for this important work Dawn Song and https://lnkd.in/g3vzFt3d. Our MLCommons AI Risk and Reliability working group is combining this type of rigorous empirical analysis from academia with input from civil society and industry--because as you note, doing this right means working together. https://lnkd.in/gkykZVkq.

Sharing MLCommons Executive Director, Rebecca Weiss's thoughts on her recent panel Credo AI's Responsible AI Leadership Summit. "As AI technologies rapidly evolve, the necessity for standardized, operationally relevant methods to evaluate their performance and safety becomes increasingly urgent. Today's conversation underscored the critical need for harmonized AI standards that recognize and adapt to the diverse requirements of different countries and regions. At MLCommons, we’re not starting from scratch. We’re drawing upon the collective expertise of our consortium – which spans engineers, scientists, and academics – to forge tools that uphold the rigor of scientific validation and are also practical for widespread enterprise application." Watch this space for additional announcements coming from our AI Risk and Responsibility working group as we are fast approaching the release of our first AI Safety benchmark, designed to evaluate the performance and safety of LLM's. Thanks for sharing more about this important work yesterday Rebecca.

Join Us at MICCAI 2024! We’re excited to invite you to our tutorial on the Comprehensive Open Federated Ecosystem (COFE) at MICCAI on October 10th at MICCAI!   Discover the latest advancements in federated learning and how COFE is revolutionizing the field. Don’t miss this opportunity to learn, network, and innovate!   More details: https://lnkd.in/g3JmdJE2 #MICCAI2024 #FederatedLearning #COFE #AI #MachineLearning Spyridon Bakas, Qi Dou, Sarthak Pati, Siddhesh Thakur, Cyril Zakka, MD, Patrick Foley, Hasan Kassem IU Division of Computational Pathology, The Chinese University of Hong Kong, Hugging Face, Intel AI

Join us today, Oct. 3rd, from 1:15 PM to 1:50 PM (PST) for an insightful panel discussion at the Credo AI-AI Summit 2024, featuring MLCommons Executive Director, Rebecca Weiss. The panel will explore how global economies approach #AI standards and the importance of harmonizing diverse models for responsible AI adoption. If you're attending, be sure to catch this engaging conversation! Learn more about the summit here: https://lnkd.in/gFiiizdJ #DataTransparency #CredoAISummit #AIRiskandResponsibility #MLCommons #AIInnovation

Join us tomorrow, Oct 3rd, 1:15-1:50 p.m. PST, at the Credo AI Summit 2024 to explore the future of responsible AI! MLCommons Executive Director Rebecca Weiss will speak on a panel about global AI standards and governance approaches. Don't miss out—register now to secure your spot: https://lnkd.in/eC56APTT #DataTransparency #CredoAISummit #AIRiskandResponsibility #MLCommons #AIInnovation

From the State of California, Governor Gavin Newsom's veto of the bill SB 1047 amplifies the ongoing debate surrounding AI safety and innovation. As policymakers and the public seek solutions that balance innovation with risk mitigation, the need for robust, standardized benchmarks is more pressing than ever. At MLCommons, we believe that transparent and standardized benchmarking systems can help pave the way toward more reliable, compliant, and safer AI models. While debates over specific legislation will persist, benchmarks provide a clear framework to evaluate AI model operations, ensuring that safety and compliance measures are met without stifling innovation. As we continue to develop our AI Risk and Reliability benchmarks, our goal is to offer tools that link operational excellence to safety and reliability—helping developers, regulators, and the broader public work together toward a safer AI future. #AIsafety #SB1047