腾讯混元2.0上线

Core Insights - Tencent's latest language models, Tencent HY 2.0 Think and Tencent HY 2.0 Instruct, have been officially released, showcasing advancements in AI technology [2] Group 1: Model Specifications - The HY 2.0 model utilizes a mixture of experts (MoE) architecture, featuring a total of 406 billion parameters and 32 billion active parameters [2] - The model supports a context window of 256,000 tokens, enhancing its capability for processing large amounts of data [2] Group 2: Improvements Over Previous Version - Compared to the previous version (Hunyuan-T1-20250822), HY 2.0 Think has significantly improved pre-training data and reinforcement learning strategies [2]