THU-KEG/SIRI-7B-high

THU-KEG/SIRI-7B-high is a 7.6 billion parameter large reasoning model (LRM) developed by THU-KEG, utilizing the SIRI (Scaling Iterative Reinforcement Learning with Interleaved Compression) framework. This model is specifically designed to enhance the efficiency and accuracy of reasoning tasks by iteratively balancing concise reasoning with exploratory planning. It achieves token efficiency without compromising accuracy, making it suitable for applications requiring optimized reasoning performance.