arXiv 2512.02556
DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
By DeepSeek-AI, Aixin Liu, et al.
Published 2025-12-02
Wiki summary
Explore the paper's summary, context, and related research on Papiers.
We introduce DeepSeek-V3.2, a model that harmonizes high computational efficiency with superior reasoning and agent performance. The key technical breakthroughs of DeepSeek-V3.2 are as follows: (1) DeepSeek Sparse Attention (DSA): We introduce DSA, an efficient attention mechanism that substantially reduces computational complexity while preserving model performance in long-context scenarios. (2) Scalable Reinforcem…