评分依据:5B model with Dual-Space Sparse Attention for long-context efficiency. Notable architecture innovation.
SpikingBrain2.0: Brain-Inspired Foundation Models for Efficient Long-Context and Cross-Platform Inference
发布
采集
学术前沿 7.0 分
— 5B model with Dual-Space Sparse Attention for long-context efficiency. Notable architecture innovation. 原文: arxiv.org