arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2505.20643cs.LGcs.AI
收藏

过往经验能否加速大语言模型推理?

Can Past Experience Accelerate LLM Reasoning?

Bo Pan;Liang Zhao

Bo Pan, Liang Zhao

AI总结:

本文研究大语言模型能否通过过往经验加速推理,提出SpeedupLLM框架,实验表明在合适记忆与推理方法下计算成本最多降低56%。

摘要:

将更多计算资源分配给大语言模型(LLM)的推理通常已被证明能提升其有效性,但同时也导致推理时间增加。相比之下,人类随着经验和接触的增加,可以更快更好地完成任务。因此,本文旨在探讨以下问题:大语言模型能否通过反复接触相关任务而变得推理更快?如果可以,又该如何实现?为回答这些问题,我们首先在任务相关性和计算预算计算两个维度上,系统性地形式化了大语言模型推理加速的问题设定。接着,我们提出了SpeedupLLM,一个具有理论保证的框架,基于自适应计算分配和记忆机制来实施并基准测试这种推理加速行为。我们进一步开展了全面的实验,在不同问题相似度水平、记忆方法和推理方法下对此行为进行基准测试。结果表明,大语言模型通常能借助过往经验更快地推理,在配备合适的记忆和推理方法时,计算成本最多可降低56%。

英文摘要:

Allocating more compute to large language models (LLMs) reasoning has generally been demonstrated to improve their effectiveness, but also results in increased inference time. In contrast, humans can perform tasks faster and better with increased experience and exposure. Hence, this paper aims to investigate the question: Can LLMs also become faster at reasoning through recurrent exposure on relevant tasks, and if so, how can it be achieved? To address these questions, we first formalize the problem setting of LLM reasoning speedup systematically in the dimensions of task relevancy and compute budget calculation. We then propose SpeedupLLM, a theoretically guaranteed framework to implement and benchmark such reasoning speedup behaviour based on adaptive compute allocation and memory mechanisms. We further conduct comprehensive experiments to benchmark such behaviour across different question similarity levels, memory methods, and reasoning methods. Results show that LLMs can generally reason faster with past experience, achieving up to a 56% reduction in compute cost when equipped with appropriate memory and reasoning methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑