Category: Data & Computing
-
Widely used AI tool for early sepsis detection may be cribbing doctors’ suspicions
When using only data collected before patients with sepsis received treatments or medical tests, the model’s accuracy was no better than a coin toss.
-
U-M team receives NIH grant for collaborative research to speed ARDS diagnosis
University of Michigan researchers examine if molecular compounds in exhaled breath could lead to improved diagnosis and tracking of acute respiratory distress syndrome (ARDS).
-
Large open dataset aims to improve understanding of building electricity demand response
Data collected from 14 commercial buildings can help inform efforts to balance electrical grids, maintaining reliability.
-
Clinicians could be fooled by biased AI, despite explanations
Regulators pinned their hopes on clinicians being able to spot flaws in explanations of an AI model’s logic, but a study suggests this isn’t a safe approach.
-
Understanding attention in large language models
How do chatbots based on the transformer architecture decide what to pay attention to in a conversation? They’ve made their own machine learning algorithms to tell them.
-
Biases in large image-text AI model favor wealthier, Western perspectives
AI model that pairs text, images performs poorly on lower-income or non-Western images, potentially increasing inequality in digital technology representation.
-
Nextgen computing: Hard-to-move quasiparticles glide up pyramid edges
Computing with a combination of light and chargeless excitons could beat heat losses and more, but excitons need new modes of transport.
-
$18M for campus-wide ecosystem for designing and manufacturing materials
New center to advance materials research for quantum computing, sustainable plastics and more, while training a more representative workforce.
-
Optimization could cut the carbon footprint of AI training by up to 75%
Deep learning models that power giants like TikTok and Amazon, as well as tools like ChatGPT, could save energy without new hardware or infrastructure.
-
Scalable method to manufacture thin film transistors achieves ultra-clean interface for high performance, low-voltage device operation
Led by Prof. Becky Peterson, the research focuses on a category of materials important for low power logic operations, high pixel density screens, touch screens, and haptic displays.
-
Six ECE faculty will help shape the future of semiconductors as part of the JUMP 2.0 program
Elaheh Ahmadi, David Blaauw, Michael Flynn, Hun-Seok Kim, Hessam Mahdavifar, and Zhengya Zhang bring their expertise and creativity to this nationwide undertaking in the area of semiconductors and information & communication technologies.
-
Open-source hardware: a growing movement to democratize IC design
Dr. Mehdi Saligane, a leader in the open-source chip design community, was among the first researchers to fabricate a successful chip as part of Google’s multi-project wafer program.