Search AI Products and News

LG AI Research recently open-sourced its EXAONE Deep Reasoning AI model, touted as globally noteworthy. Its biggest highlight is its reasoning capabilities. Simply put, it can independently formulate hypotheses and verify them through reasoning, enabling autonomous decision-making. This marks a new era of Agentic AI, where AI may truly begin to develop its own intelligence. LG AI Research proudly announced the...

iFlytek has announced a significant upgrade to its deep reasoning model, StarFire X1. This upgrade focuses on enhancing its capabilities across various domains, including mathematics, healthcare, and education, delivering notable improvements and innovations. StarFire X1 exhibits a significant enhancement in mathematical capabilities, particularly in tackling competition-level problems. Compared to other industry models, StarFire X1 demonstrates competitive performance in mathematical tasks while maintaining a smaller model parameter count, rivaling models like DeepSeek R1 and comparable OpenAI models. Users can access StarFire through iFlytek's...

Training the deep learning model RECAST with a larger dataset provides higher flexibility compared to existing models. RECAST is based on neural network generative models and can predict the timing of the next earthquake based on historical data. Utilizing GPU acceleration for training, RECAST has shown significant improvements in accuracy across multiple regions compared to existing models. The researchers plan to open-source this model for more teams to test and iterate for improvements.

The research team at Daegu University of Science and Technology in South Korea has successfully developed a small sample learning model that can accurately classify using only a small amount of brainwave data, representing a revolutionary advancement. This model overcomes the limitation of traditional deep learning models, which require large volumes of data, and is expected to drive new breakthroughs in brainwave-related research. The research team utilized embedding modules, temporal attention modules, aggregation attention modules, and relational modules to enhance the classification accuracy of the model. The new deep learning model demonstrates up to 76% classification accuracy in cross-subject classification, far surpassing previous methods. This research has promising applications in medical fields.