InfLLM-V2: Dense-Sparse Switchable Attention for Seamless Short-to-Long Adaptation Paper • 2509.24663 • Published Sep 29 • 13
APB: Accelerating Distributed Long-Context Inference by Passing Compressed Context Blocks across GPUs Paper • 2502.12085 • Published Feb 17 • 4
FR-Spec: Accelerating Large-Vocabulary Language Models via Frequency-Ranked Speculative Sampling Paper • 2502.14856 • Published Feb 20 • 8