
Alibaba's Latest AI Model, QwQ-32B, Beats Larger Rivals in Math and Reasoning
Despite having far fewer parameters, Qwen’s new QwQ-32B model outperforms DeepSeek-R1 and OpenAI’s o1-mini in mathematical benchmarks and scientific reasoning, showcasing the power of reinforcement learning.