Efficient Beam Search for Large Language Models Using Trie-Based Decoding

This work presents a novel trie (prefix-tree)based parallel decoding method that addresses the memory inefficiency of batch-based beam search. By sharing a single KV cache across beams with common prefixes, our approach dramatically reduces memory usage and enables efficient decoding. We evaluated our method across three attention architectures, Multi-Head Attention (Phi-3.5-miniinstruct), Grouped Query Attention (Llama-3.1-8B-Instruct), and Sliding Window Attention (Mistral-Small-24B-Instruct-2501), using CN-N/DailyMail for abstractive summarization and HumanEval for code generation. Our experiments demonstrate substantial memory savings (4-8×) and up to 2.4× faster decoding, without compromising generation quality. These results highlight our method's suitability for memory-constrained environments and largescale deployments.