1
AI Is Designing Radio Chips That Humans Couldn’t Even Imagine
Princeton researchers are using reinforcement learning and diffusion models to automate and accelerate the design of radio frequency integrated circuits.
Top stories in Research
1
David Autor named head of the Department of Economics
MIT has appointed economist David Autor, a prominent researcher on AI and the future of labor, as the head of its Department of Economics.
1
Introducing the FFASR Leaderboard: Benchmarking ASR in the Real World
Hugging Face has launched the FFASR Leaderboard to benchmark automatic speech recognition performance in real-world scenarios.
1
New chip could help tiny robots traverse complex environments
MIT researchers have designed a low-power chip and algorithm that enables small robots to construct 3D navigation maps using minimal memory.
1
AI Is Learning to Read the Room
Researchers are developing AI models capable of analyzing subtle physical and vocal cues to better understand human emotional states and context.
0
Using AI to help physicians diagnose rare genetic diseases affecting children
Researchers successfully used an OpenAI reasoning model to identify eighteen new diagnoses in previously unsolved cases of rare genetic pediatric diseases.
0
A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry
OpenAI and Molecule.one demonstrated that a near-autonomous AI system can optimize complex chemical reactions to assist in medicinal chemistry and drug discovery.
0
A better way to model the behavior of metal alloys
MIT researchers have developed an AI-driven modeling approach that improves predictions of metal alloy properties by capturing subtle atomic patterns.
0
MosaicLeaks: Can your research agent keep a secret?
A new study on Hugging Face examines the security risks of research agents inadvertently leaking sensitive information.
0
Introducing LifeSciBench
OpenAI has introduced LifeSciBench, an expert-reviewed benchmark designed to evaluate how effectively AI systems perform real-world life science research tasks.
0
Sound Waves Give Neuromorphic Chips a Brain-Simulating Edge
A new study demonstrates that utilizing sound waves in neuromorphic devices can improve their energy efficiency and processing speed compared to electronic alternatives.
0
Is it agentic enough? Benchmarking open models on your own tooling
Hugging Face has published a guide on benchmarking open-source models to evaluate their agentic capabilities using custom tools.
0
Securing the future of AI agents
Google DeepMind has proposed an AI Control Roadmap that combines traditional safeguards with real-time monitoring to secure internal systems against potential risks from autonomous agents.
0
Beyond LoRA: Can you beat the most popular fine-tuning technique?
Hugging Face explored alternative fine-tuning methodologies that aim to surpass the performance of the widely used Low-Rank Adaptation technique.
0
Predicting model behavior before release by simulating deployment
OpenAI has introduced Deployment Simulation, a new evaluation method that uses real conversation data to predict AI model behavior and safety risks prior to release.
0
Quantum Pulse Does Industrial Light Magic To Deliver Massive Boost In Qubits
Quantum Pulse has introduced a new method to significantly increase qubit capacity in quantum computing systems.
0
From pixels to planning: Earth AI for nature restoration
Google Research has introduced Earth AI tools designed to assist in nature restoration and environmental planning by analyzing geospatial data.
0
Agentic Resource Discovery: Let agents search
Hugging Face explored the concept of agentic resource discovery, focusing on enabling AI agents to autonomously search for and locate digital resources.
0
In game theory, generalists sometimes win out over specialists
Researchers have demonstrated that a previously overlooked class of generalist algorithms can outperform specialist algorithms in certain game theory scenarios.
0
Could AI tell you where you left your keys?
Researchers have developed a spatial memory system that helps robots efficiently track and recall the locations of physical objects in their environment.
0
MIT’s Initiative for New Manufacturing builds momentum
MIT's Initiative for New Manufacturing completed its first year of advancing manufacturing technologies through research, workforce training, and industry collaboration.
0
Ire identifies another LOTUSLITE specimen
Microsoft's Project Ire has successfully identified and reverse-engineered a new specimen of the LOTUSLITE malware that went undetected by major security tools.
0
A low-carbon computing platform from your retired phones
Google Research has proposed a low-carbon computing platform that repurposes retired smartphones to handle distributed computational tasks.
0
Visual Language Models Train Robots to Read Human Emotions
Researchers have utilized visual language models to train collaborative robots to interpret human emotions by analyzing facial expressions and contextual cues.