Transformers have taken the machine learning world by storm with their powerful self-attention mechanism, achieving state-of-the-art results in areas like natural language processing and computer vision. However, when it came to graph data, which is ubiquitous in domains such as social networks, biology, and chemistry, the classic Transformer models hit a major bottleneck due to… →

IBM has made a great advancement in the field of software development by releasing a set of open-source Granite code models designed to make coding easier for people everywhere. This action stems from the realization that, although software plays a critical role in contemporary society, the process of coding is still difficult and time-consuming. Even… →

In Natural Language Processing (NLP) tasks, data cleaning is an essential step before tokenization, particularly when working with text data that contains unusual word separations such as underscores, slashes, or other symbols in place of spaces. Since common tokenizers frequently rely on spaces to split text into distinct tokens, this problem can have a major… →

In cybersecurity, while AI technologies have significantly bolstered our defense mechanisms against cyber threats, they have also given rise to a new era of sophisticated attacks. Let’s explore the darker side of AI advancements in the cybersecurity domain, focusing on its role in enhancing adversarial capabilities. From AI-powered phishing attacks that craft deceptively personal messages… →

The challenge of training large and sophisticated models is significant, primarily due to the extensive computational resources and time these processes require. This is particularly evident in training large-scale Generative AI models, which are prone to frequent instabilities manifesting as disruptive loss spikes during extended training sessions. Such instabilities often lead to costly interruptions that… →

Recently, there’s been increasing interest in enhancing deep networks’ generalization by regulating loss landscape sharpness. Sharpness Aware Minimization (SAM) has gained popularity for its superior performance on various benchmarks, specifically in managing random label noise, outperforming SGD by significant margins. SAM’s robustness shines particularly in scenarios with label noise, showcasing substantial improvements over existing techniques.… →

Rightsify’s Global Copyright Exchange (GCX) offers vast collections of copyright-cleared music datasets tailored for machine learning and generative AI music initiatives.  These datasets encompass millions of hours of music, over 10 million recordings and compositions accompanied by comprehensive metadata, including key, tempo, instrumentation, keywords, moods, energies, chords, and more, facilitating training and commercial usage.  Text,… →