@article{positionawaredepthdecaydecodingd3, title = {Position-Aware Depth Decay Decoding ($D^3$): Boosting Large Language Model Inference Efficiency}, author = {Siqi Fan and Xuezhi Fang and Xingrun Xing and Peng Han and Shuo Shang and Yequan Wang}, year = {2025}, eprint = {2503.08524}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2503.08524v1}, }