SenseTime Technology has officially launched the "Daily New" Fusion Large Model. This model has achieved significant improvements in multimodal information processing and deep reasoning capabilities, securing the top position in two authoritative evaluation rankings.
According to the "2024 Annual Report on Chinese Large Model Benchmark Evaluation" released by the domestic authoritative evaluation agency SuperCLUE, SenseTime's "Daily New" Fusion Large Model shares the top spot in the country with a high score of 68.3 alongside DeepSeek V3. Additionally, in the multimodal evaluation by OpenCompass, this model also ranked first, significantly surpassing GPT-4o.
The "Daily New" Fusion Large Model from SenseTime marks a substantial breakthrough in the field of native multimodal training. This means the model can integrate various types of information, such as text, images, and videos, bringing new development directions to the industry. In the SuperCLUE evaluation, the model excelled in both humanities and science scores, achieving a global first place in humanities with a score of 81.8, while also winning gold in science, where it ranked first in the country with a score of 78.2 in the computational dimension.
The "Daily New" Fusion Large Model is characterized by its ability to simultaneously handle various complex pieces of information. It can not only recognize vague text but also quickly analyze data charts and provide in-depth reasoning support. In practical application scenarios, this model has demonstrated its unique advantages, particularly in fields such as autonomous driving, video interaction, office education, finance, and industrial manufacturing.
During the product development process, SenseTime employed a large amount of data with mixed text and images, establishing interactive bridges between modalities through advanced synthesis technology, enhancing the model's understanding of multimodal information. Furthermore, SenseTime has built a rich set of cross-modal tasks, providing a solid foundation for training the "Daily New" Fusion Large Model. Through this innovative training method, SenseTime's fusion modal model can effectively respond to user needs across various business scenarios, achieving a positive cycle of application implementation and foundational model iteration.
Key Points:
🌟 SenseTime Technology has launched the "Daily New" Fusion Large Model, showcasing outstanding performance.
📊 The model excels in both humanities and science scores, ranking first globally in humanities and first domestically in science.
🚀 The model is applicable in multiple fields such as autonomous driving, finance, and online education, demonstrating powerful multimodal processing capabilities.
