Can AI grasp related concepts after learning only one?
Researchers have now developed a technique that advances the ability of these tools, such as ChatGPT, to make compositional generalizations. This technique, Meta-learning for Compositionality, outperforms existing approaches and is on par with, and in some cases better than, human performance.
Powerful machine-learning models are being used to help people tackle tough problems such as identifying disease in medical images or detecting road obstacles for autonomous vehicles. But machine-learning models can make mistakes, so in high-stakes settings it's critical that humans know when to trust a model's predictions.
As the perpetrators of accounting fraud become ever more sophisticated in their techniques, fraud detection needs to step up its game. Thankfully, a group of researchers have devised a new machine learning 'detective' that is able to analyze not just fraud at a single firm, but predict likely fraud across…
Training a large artificial intelligence model is expensive, not just in dollars, but in time, energy, and computational resources. Traditionally, obtaining a smaller, faster model either requires training a massive one first and then trimming it down, or training a small one from scratch and accepting weaker performance.