DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation Paper • 2607.05147 • Published Jul 6 • 41
Single-Rollout Asynchronous Optimization for Agentic Reinforcement Learning Paper • 2607.07508 • Published Jul 8 • 29
DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence Paper • 2606.19348 • Published Apr 26 • 28