上下文学习是一种误差驱动学习吗？来自结构启动中反频率效应的证据
Is In-Context Learning a Type of Error-Driven Learning? Evidence from the Inverse Frequency Effect in Structural Priming
AI总结：
本文通过结构启动中的反频率效应诊断上下文学习，发现大语言模型表现出该效应且随模型规模增强，证明ICL是一种误差驱动学习。
摘要：
大型语言模型（LLMs）已展现出上下文学习（ICL）的涌现能力。一类研究声称，ICL在功能上等价于梯度下降，这是一种误差驱动学习机制。在本文中，我们引入一种新的诊断方法，用于判断ICL是否在功能上执行误差驱动学习。我们的方法基于反频率效应（IFE）——一种现象，即当智能体面对不太可能出现的示例时，其行为受到的影响程度要大于面对更可能出现的示例时。IFE此前已在心理语言学中被识别出来，人类在结构启动（人们倾向于产出最近遇到过的句子结构）的语境中表现出IFE。在该语境中，IFE已被用作证据，表明人类结构启动必然涉及误差驱动学习机制。在我们的实验中，我们用ICL模拟了结构启动，并发现LLMs确实表现出IFE，且该效应在更大的模型中更强。我们得出结论，至少在我们研究的案例中，ICL确实是一种误差驱动学习，这支持了在ICL期间前向传播中隐式计算误差信号的假设。我们的结果表明，人类和LLMs在在线处理中都利用了误差驱动处理机制。
英文摘要：
Large language models (LLMs) have shown the emergent capability of in-context learning (ICL). One line of research has claimed that ICL is functionally equivalent to gradient descent, a type of error-driven learning mechanism. In this paper, we introduce a new way of diagnosing whether ICL is functionally performing error-driven learning. Our approach is based on the inverse frequency effect (IFE) -- a phenomenon in which an agent's behavior is influenced to a greater degree when presented with improbable examples as compared to more likely ones. The IFE has previously been identified in psycholinguistics where humans exhibit the IFE in the context of structural priming (the tendency for people to produce sentence structures they have encountered recently). In that context, the IFE has been used as evidence that human structural priming must involve error-driven learning mechanisms. In our experiments, we simulated structural priming with ICL and found that LLMs indeed display the IFE, with the effect being stronger in larger models. We conclude that at least in the case we studied, ICL is indeed a type of error-driven learning, supporting the hypothesis that an error signal is implicitly computed in the forward pass during ICL. Our results suggest that both humans and LLMs make use of error-driven processing mechanisms in on-line processing.
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。
