An international team of scientists has used machine learning to help them develop perovskite solar cells with near-record efficiency. In their paper published in the journal Science, the group describes how they used the machine-learning algorithm to help them find new hole-transporting materials to improve the efficiency of perovskite solar cells.
Accelerating LLM inference is an important ML research problem, as auto-regressive token generation is computationally expensive and relatively slow, and improving inference efficiency can reduce latency for users. In addition to ongoing efforts to accelerate inference on Apple silicon, we have recently made significant progress in accelerating LLM inference for the NVIDIA GPUs widely used …
Read more “Accelerating LLM Inference on NVIDIA GPUs with ReDrafter”
This post is co-written with Marta Cavalleri and Giovanni Germani from Fastweb, and Claudia Sacco and Andrea Policarpi from BIP xTech. AI’s transformative impact extends throughout the modern business landscape, with telecommunications emerging as a key area of innovation. Fastweb, one of Italy’s leading telecommunications operators, recognized the immense potential of AI technologies early on …
Retrieval-augmented generation (RAG) supercharges large language models (LLMs) by connecting them to real-time, proprietary, and specialized data. This helps LLMs deliver more accurate, relevant, and contextually aware responses, minimizing hallucinations and building trust in AI applications. But RAG can be a double-edged sword: while the concept is straightforward – find relevant information and feed it …
Barry Wilmore and Suni Williams will now come home in March at the earliest, to allow SpaceX and NASA engineers to complete development of a new Crew Dragon spacecraft.
Scientists have developed swarms of tiny magnetic robots that work together like ants to achieve Herculean feats, including traversing and picking up objects many times their size. The findings suggest that these microrobot swarms — operating under a rotating magnetic field — could be used to take on difficult tasks in challenging environments that individual …
Read more “Swarms of ‘ant-like’ robots lift heavy objects and hurl themselves over obstacles”
In a new study, participants tended to assign greater blame to artificial intelligences (AIs) involved in real-world moral transgressions when they perceived the AIs as having more human-like minds. Minjoo Joo of Sookmyung Women’s University in Seoul, Korea, presents these findings in the open-access journal PLOS ONE on December 18, 2024.
The adoption of machine learning (ML) continues at a rapid pace, as it has proven itself a powerful tool for solving many problems.
Our comprehensive benchmark and online leaderboard offer a much-needed measure of how accurately LLMs ground their responses in provided source material and avoid hallucinations
Teleoperation for robot imitation learning is bottlenecked by hardware availability. Can high-quality robot data be collected without a physical robot? We present a system for augmenting Apple Vision Pro with real-time virtual robot feedback. By providing users with an intuitive understanding of how their actions translate to robot motions, we enable the collection of natural …
Read more “ARMADA: Augmented Reality for Robot Manipulation and Robot-Free Data Acquisition”