Artificial intelligence, machine learning, computer science and mathematics.
Researchers tested a programmable hardware design that imitates selected quantum-computing ideas to run matrix calculations in parallel. In representative benchmarks, the design reached up to 6.3 times the speed and 4.1 times the energy efficiency of conventional matrix co-processors and embedded graphics processors.
Researchers evaluated three AI systems on hundreds of molecules bound to a SARS-CoV-2 protein and in three prospective drug screens. The results suggest that combining AI-based predictions with traditional docking may help prioritize candidates, although no single approach worked best in every test.
Researchers developed a hardware-assisted system intended to detect message injection, spoofing, flooding and impersonation attacks on automotive networks. In tests using simulated traffic and a prototype automotive processor system, it reported an average detection time of 14 microseconds and less than 2 percent CPU overhead.
An exploratory study compared four ways to develop software with and without conversational AI systems. Human–AI collaboration produced higher-quality, better-organized test suites, while autonomous AI agents generated functionally correct code more quickly but added complexity and untested decisions.
Researchers developed an AI system that transfers camera movement from one video to a user’s still image. In studies with 72 people, participants more often preferred its motion and scene preservation over two comparison methods.
An AI system for writing formal mathematical proofs solved nine of 353 open problems associated with mathematician Paul Erdős. It also proved 44 of 492 conjectures from the Online Encyclopedia of Integer Sequences, with each accepted proof checked by the Lean proof system.
A qualitative study of 24 university students found that neurodivergent participants used generative AI to manage academic demands, energy and cognitive processing. They described it as essential scaffolding but also raised concerns about authenticity, over-reliance and interfaces that may exclude some users.
A lightweight security pairing for low-power internet devices used 52% less memory and 37% less energy than AES-CCM in tests. The researchers also tested it against differential, linear and replay attacks on two microcontrollers.
A systematic survey examines how large language model-based AI agents are being developed to resolve software issues described in everyday language. It maps the field’s benchmarks, techniques and studies, and identifies reinforcement learning as an increasingly important training approach.
A new voice-obfuscation system runs on a $3 microcontroller between a microphone and a smart speaker. The system is designed to reduce voice-based identification before audio leaves the device while preserving the speaker’s ability to understand and respond to commands.
A system called ARC converted large requirement documents containing hundreds of scenarios into runnable web and Android software. Across two benchmarks, the systems passed more automated interface tests than other tested AI-based programming systems, while also producing designs, tests and links between requirements and code.
A system called IntentionTest generates software tests from a developer’s description of what a function should do in a particular situation. Across 3,680 test cases from 12 open-source projects, it outperformed four other test-generation methods on measures of relevance, coverage and successful execution.
Researchers introduced a benchmark and an AI-based framework for testing and improving the security of generated Python code. Across multiple language-model systems, the framework improved both first-attempt functional success and resistance to targeted security tests compared with existing approaches.
A study of 13 researchers informed an AI-augmented dashboard for monitoring long-running passive sensing studies. Tests of the prototype found promise for daily monitoring and decision-making, while highlighting concerns about transparency and researchers’ expectations of AI.
Researchers tested three Java code-checking tools and identified 14 previously unreported false negatives or false positives. Developers confirmed 11 of the issues, and nine had been fixed at the time of the study.
