Qin Z, He Z, Prakriya N, Cong J, Sun Y. Dynamic-Width Speculative Beam Decoding for LLM Inference. AAAI [Internet]. 2025 Apr. 11 [cited 2026 May 25];39(23):25056-64. Available from: https://ojs.aaai.org/index.php/AAAI/article/view/34690