哪种形式的照护者反馈支持语法学习？一项关于类儿童语言模型的强化学习研究
Which Forms of Caregiver Feedback Support Grammar Learning? A Reinforcement-Learning Study of Child-Like Language Models
- ENS, Université PSL, EHESS, CNRS（巴黎高等师范学院、巴黎文理研究大学、高等社会科学研究学院、法国国家科学研究中心）
- Aix Marseille Université, CNRS（艾克斯-马赛大学、法国国家科学研究中心）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本研究用类儿童语言模型测试四种照护者反馈对语法学习的影响，发现结构对齐最有效，交际反馈次之，语义关联和情感反馈无益于语法。
AI中文摘要：
社会互动是儿童语言学习的核心，但在自然数据中，不同形式的照护者反馈的效果难以分离。我们使用类儿童语言模型作为受控学习者，测试哪种形式的反馈支持语法发展。小型GPT-2风格模型在CHILDES的儿童导向语言上预训练，然后使用奖励模型进行强化学习微调，奖励模型训练用于捕捉四种反馈类型：交际反馈、结构对齐、语义关联和情感反馈。奖励微调在最小对评估上产生有限收益，但在自由生成中效果更明显。结构对齐在语法性方面产生最强改进，为这种反馈如何支持语法学习提供了一种新颖、合理的机制性解释。交际反馈产生更适度的收益。相比之下，语义关联和情感反馈不改善语法性，尽管进一步分析表明它们可能支持语法之外的语言学习的其他方面。这些结果表明，不同形式的照护者反馈对语言学习做出互补贡献。
英文摘要：
Social interaction is central to children's language learning, but the effects of different forms of caregiver feedback are difficult to isolate in naturalistic data. We use child-like language models as controlled learners to test which forms of feedback support grammatical development. Small GPT-2-style models are pretrained on child-directed language from CHILDES, then fine-tuned with reinforcement learning using reward models trained to capture four feedback types: communicative feedback, structural alignment, semantic contingency, and affective feedback. Reward fine-tuning yields limited gains on minimal-pair evaluations, but clearer effects in free generation. Structural alignment produces the strongest improvements in grammaticality, providing a novel, plausible mechanistic account of how this feedback can support grammar learning. Communicative feedback yields more moderate gains. In contrast, semantic contingency and affective feedback do not improve grammaticality, although further analyses suggest that they may support other aspects of language learning beyond grammar. These results suggest that different forms of caregiver feedback make complementary contributions to language learning.
