Inthis paper,we present a neural network that cansimultaneously learn the dynamics of the system and the constants of motion from data.

We will incorporate this feedback.

On 28 March, Briony Swire-Thompson began seeing reports online that the National Institutes of Health (NIH) might cancel grants for research on misinformation. At first, she didn't think she would be affected. Swire-Thompson, a psychologist at Northeastern University, studies misinformation--but not the political lies that get most of the attention. She's interested in false information about cancer, and why people fall for it. "There's a lot of people online trying to sell their snake oil," she says.

The current U.S. administration has launched a wara on academia. Indirect costs, or, more accurately, facility and administration expenses, support research but cannot be directly attributed to a specific project, such as lab infrastructure, utilities, and administrative support. These are real costs; the limit, which has since been suspended by courts, is a severe blow to biomedical research in the U.S. Beyond expanding this limit to other agencies, such as the National Science Foundation (NSF), the administration is also reportedly considering slashing NSF's annual budget from approximately US 9 billion down to about US 3– 4 billion. This would deal a devastating blow to academic U.S. research, especially computing research. As statedc by the Computing Research Association (CRA), "NSF budget cuts would put the future of U.S. innovation and security at risk."

Nearly three months into President Donald Trump's term, the future of American AI leadership is in jeopardy. Basically any generative-AI product you have used or heard of--ChatGPT, Claude, AlphaFold, Sora--depends on academic work or was built by university-trained researchers in the industry, and frequently both. Today's AI boom is fueled by the use of specialized computer-graphics chips to run AI models--a technique pioneered by researchers at Stanford who received funding from the Department of Defense. They rely on a training method called "reinforcement learning," the foundations of which were developed with National Science Foundation (NSF) grants. "I don't think anybody would seriously claim that these [AI breakthroughs] could have been done if the research universities in the U.S. didn't exist at the same scale," Rayid Ghani, a machine-learning researcher at Carnegie Mellon University, told me.

We address the challenge of federated learning on graph-structured data distributed across multiple clients. Specifically, we focus on the prevalent scenario of interconnected subgraphs, where interconnections between different clients play a critical role. We present a novel framework for this scenario, named FedStruct, that harnesses deep structural dependencies. To uphold privacy, unlike existing methods, FedStruct eliminates the necessity of sharing or generating sensitive node features or embeddings among clients. Instead, it leverages explicit global graph structure information to capture inter-node dependencies. We validate the effectiveness of FedStruct through experimental results conducted on six datasets for semi-supervised node classification, showcasing performance close to the centralized approach across various scenarios, including different data partitioning methods, varying levels of label availability, and number of clients.

The government's campaign to fight "misinformation" has expanded to adapt military-grade artificial intelligence once used to silence the Islamic State (ISIS) to quickly identify and censor American dissent on issues like vaccine safety and election integrity, according to grant documents and cyber experts. The National Science Foundation (NSF) has awarded several million dollars in grants recently to universities and private firms to develop tools eerily similar to those developed in 2011 by the Defense Advanced Research Projects Agency (DARPA) in its Social Media in Strategic Communication (SMISC) program. DARPA said those tools were used "to help identify misinformation or deception campaigns and counter them with truthful information," beginning with the Arab Spring uprisings in the the Middle East that spawned ISIS over a decade ago. The initial idea was to track dissidents who were interested in toppling U.S.-friendly regimes or to follow any potentially radical threats by examining political posts on Big Tech platforms. Mike Benz, executive director of the Foundation for Freedom Online has compiled a report detailing how this technology is being developed to manipulate the speech of Americans via the National Science Foundation (NSF) and other organizations.

ExLattice, Inc, a manufacturing AI company, has received a Phase I award from the National Science Foundation (NSF) Small Business Innovation Research (SBIR) Program for developing its accelerated simulation engine dedicated to additive manufacturing. The Phase I SBIR grant, valued at over $250,000, will be used to develop and validate ultrafast manufacturing simulation solutions in collaboration with multiple universities. The goal is to cut the time-consuming steps in computation and deliver real-time engineering solutions for users to understand, control and improve additive manufacturing systems and outcomes. "Receiving the SBIR award from NSF is another proof of our vision in engineering software for digital manufacturing," says Dr. Runze Huang, CEO, ExLattice. "The NSF SBIR grant not only provides us the resources, but a platform to collaborate with leading experts in academia and great business partners in bringing AI to manufacturing."

Much of the focus on the CHIPS and Science Act of 2022 has centered around strengthening the semiconductor industry, but the lion's share of the funds being allotted in the $280 billion bill targets industry and academic institutions conducting research and development on emerging technologies. The federal government will dole out around $50 billion over the next five years, mainly to chip manufacturers that will come out of the newly established CHIPS for America Fund. That leaves nearly $230 billion for vendors, researchers and developing technologies, and to create programs to educate students and train employees on technologies destined to become mainstream in the years ahead. Two such technologies prominently mentioned are quantum computing and AI. For instance, the National Science Foundation (NSF), one of the federal agencies responsible for allocating the funds, will establish a quantum education pilot program to promote a quantum information science workforce across the U.S. The agency also must produce a study on the educational challenges in creating a diverse and sustainable workforce in the quantum industry. NSF will also expand the existing Quantum User Expansion for Science and Technology program, making it easier for researchers to access government quantum computing hardware and clouds.

Most of the public discourse around artificial intelligence (AI) policy focuses on one of two perspectives: how the government can support AI innovation, and how the government can deter its harmful or negligent use. Yet there can also be a role for government in making it easier to use AI beneficially--in this niche, the National Science Foundation (NSF) has found a way to contribute. Through a grant-making program called Fairness in Artificial Intelligence (FAI), the NSF is providing $20 million in funding to researchers working on difficult ethical problems in AI. The program, a collaboration with Amazon, has now funded 21 projects in its first two years, with an open call for applications in its third and final year. This is an important endeavor, furthering a trend of federal support for the responsible advancement of technology, and the NSF should continue this important line of funding for ethical AI.