By Zainab Shafiq 
SenseTime has unveiled its latest AI model, SenseNova 5.0, aiming to surpass OpenAI’s GPT-4 in the race for large language model dominance.
SenseNova 5.0 boasts advancements in logical reasoning, creative writing, and human-like text generation. This innovative model combines transformer and recurrent neural network architectures, trained on a massive dataset exceeding 10 billion tokens.
A key differentiator is its ‘Mixture of Experts’ approach, allowing SenseNova 5.0 to manage a context window of approximately 200,000 tokens during inference, outperforming GPT-4 Turbo. Additionally, the model incorporates over 10TB of training data, including a significant amount of synthetic data, further refining its capabilities.
Dr. Xu Li, Chairman and CEO of SenseTime, emphasizes the ever-evolving nature of AI advancements, driven by the interplay of data, algorithms, and computing power. The integration of multi-modalities and the exponential growth in model parameters and data volume necessitate a continuous surge in computing power, shaping the future of Artificial General Intelligence (AGI).
He added,
“In the era of AGI, the three elements of data, algorithms, and computing power are undergoing a new evolution. The number of model parameters will increase exponentially, and the volume of data will grow massively with the introduction of multi modalities, leading to a continuous surge in demand for computing power.”
Also, see:
Google for Education and Pakistan Government Partner to Bridge Educational Gap