Year: 2026
Type: article
Abstract: Large language models (LLMs) and large vision-language models (LVLMs) have demonstrated impressive generative capabilities, yet ensuring their outputs align with user intent is still challenging. While most existing approaches address this issue at the training stage, inference-time approaches like decoding methods offer a more efficient and scalable solution. Decoding methods control model generation by guiding token-level selection, performing sequencelevel generation, or generating tokens in parallel to accelerate the process. In this survey, we identify three emerging paradigms from recent works on decoding methods for LLMs and LVLMs, provide a systematic review of these methods, highlight ongoing challenges, and discuss potential future research directions. Our goal is to underscore the efficiency and effectiveness of decoding methods and offer a practical view of their applications. Paper lists and more resources on decoding methods for LLMs and LVLMs can be found at https://github.com/wang2226/Awesome-LLM-Decoding.
Authors: Haoran Wang, Xiongxiao Xu, Philip S. Yu, Kai Shu
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: NSF POSE: Phase II: OpenAD: An Integrated Open-Source Ecosystem for Anomaly Detection, Collaborative Research: SaTC: CORE: Small: Targeting Challenges in Computational Disinformation Research to Enhance Attribution, Detection, and Explanation, III: Medium: Collaborative Research: Self-Supervised Recommender System Learning with Application Specific Adaption, CAREER: Towards Fairness in the Real World under Generalization, Privacy and Robustness Challenges
OpenAlex ID: W7165026022