XL3M: A Training-free Framework for LLM Length Extension Based on Segment-wise Inference

Re-assign community