# Which Forms of Caregiver Feedback Support Grammar Learning? A Reinforcement-Learning Study of Child-Like Language Models

arXiv: https://arxiv.org/abs/2609.08576 · HTML: https://mlsanity.com/paper/2609.08576/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reward modeling and RLHF, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 14 of its 56 references that are in the map.
- First posted on arXiv 8 Sep 2026.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Reward modeling and RLHF](https://mlsanity.com/alignment/reward-modeling-and-rlhf/)

## Builds on (most cited first, 14 of 14)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing](https://mlsanity.com/paper/2111.09543/) (2021)
- [BLiMP: The Benchmark of Linguistic Minimal Pairs for English](https://mlsanity.com/paper/1912.00582/) (2019)
- [A Simple Recipe for Multilingual Grammatical Error Correction](https://mlsanity.com/paper/2106.03830/) (2021)
- [Language Models Learn Rare Phenomena from Less Rare Phenomena: The Case of the Missing AANNs](https://mlsanity.com/paper/2403.19827/) (2024)
- [Evaluating Neural Language Models as Cognitive Models of Language Acquisition](https://mlsanity.com/paper/2310.20093/) (2023)
- [Babysit A Language Model From Scratch: Interactive Language Learning by Trials and Demonstrations](https://mlsanity.com/paper/2405.13828/) (2024)
- [Once Upon a Time: Interactive Learning for Storytelling with Small Language Models](https://mlsanity.com/paper/2509.15714/) (2025)
- [Dialogue Is Not Enough to Make a Communicative BabyLM (But Neither Is Developmentally Inspired Reinforcement Learning)](https://mlsanity.com/paper/2510.20358/) (2025)
- [Linguistic Generalizations are not Rules: Impacts on Evaluation of LMs](https://mlsanity.com/paper/2502.13195/) (2025)
- [BabyStories: Can Reinforcement Learning Teach Baby Language Models to Write Better Stories?](https://mlsanity.com/paper/2310.16681/) (2023)
- [Automatic Annotation of Grammaticality in Child-Caregiver Conversations](https://mlsanity.com/paper/2403.14208/) (2024)
- [Towards Developmentally Plausible Rewards: Communicative Success as a Learning Signal for Interactive Language Models](https://mlsanity.com/paper/2505.05970/) (2025)

## Abstract (arXiv)

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.