Self-Contrast:通过不一致的求解视角实现更好的反思
Self-Contrast: Better Reflection Through Inconsistent Solving Perspectives
AI总结:
针对 LLM 自我反思不稳定、反馈质量差的问题,本文提出 Self-Contrast 方法,通过探索多样化求解视角、对比差异并形成检查清单来改进反思,实验表明其在推理和翻译任务上有效且通用。
摘要:
大语言模型(LLM)的反思能力已受到广泛关注。诸如 reflexion 和 self-refine 等事后提示策略,会基于自我评估或外部反馈来改进 LLM 的响应。然而,近期研究表明,在没有外部反馈的情况下,LLM 的内在反思并不稳定。我们的研究揭示,关键瓶颈在于自我评估反馈的质量。我们发现 LLM 在自我评估时常常表现出过度自信或高度随机性,从而给出固执或不一致的反馈,导致反思效果不佳。为解决这一问题,我们提出 Self-Contrast:它根据请求自适应地探索多样化的求解视角,对比其中的差异,并将这些差异总结为一份检查清单,用于重新审视并消除不一致之处。我们的方法赋予 LLM 多样化的视角,以缓解固执偏见。此外,这些差异还揭示了 LLM 常常忽略的潜在错误或内在不确定性。对这些差异进行反思,可以促成更准确、更稳定的反思。我们在一系列推理和翻译任务上使用不同 LLM 进行的实验,证明了该策略的有效性和通用性。
英文摘要:
The reflection capacity of Large Language Model (LLM) has garnered extensive attention. A post-hoc prompting strategy, e.g., reflexion and self-refine, refines LLM's response based on self-evaluated or external feedback. However, recent research indicates without external feedback, LLM's intrinsic reflection is unstable. Our investigation unveils that the key bottleneck is the quality of the self-evaluated feedback. We find LLMs often exhibit overconfidence or high randomness when self-evaluate, offering stubborn or inconsistent feedback, which causes poor reflection. To remedy this, we advocate Self-Contrast: It adaptively explores diverse solving perspectives tailored to the request, contrasts the differences, and summarizes these discrepancies into a checklist which could be used to re-examine and eliminate discrepancies. Our method endows LLM with diverse perspectives to alleviate stubborn biases. Moreover, their discrepancies indicate potential errors or inherent uncertainties that LLM often overlooks. Reflecting upon these can catalyze more accurate and stable reflection. Experiments conducted on a series of reasoning and translation tasks with different LLMs serve to underscore the effectiveness and generality of our strategy.
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。