mlsanityData to 14 Oct 2026

Teacher Demonstrations in a BabyLM's Zone of Proximal Development for Contingent Multi-Turn Interaction

arXiv 2510.20411 ↗First posted 23 Oct 2025cs.CLcited by 9

Not on the map. Why?

Abstract (arXiv)

Multi-turn dialogues between a child and a caregiver are characterized by a property called contingency - that is, prompt, direct, and meaningful exchanges between interlocutors. We introduce ContingentChat, a teacher-student framework that benchmarks and improves multi-turn contingency in a BabyLM trained on 100M words. Using a novel alignment dataset for post-training, BabyLM generates responses that are more grammatical and cohesive. Experiments with adaptive teacher decoding strategies show limited additional gains. ContingentChat demonstrates the benefits of targeted post-training for dialogue quality and indicates that contingency remains a challenging goal for BabyLMs.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 26 of its 72 references that are in the map. First posted on arXiv 23 Oct 2025.

Builds on

26 of 72 references.

References · 26

  1. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,512
  2. 2019DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generationcited by 842
  3. 2019BLiMP: The Benchmark of Linguistic Minimal Pairs for Englishcited by 564
  4. 2024ORPO: Monolithic Preference Optimization without Reference Modelcited by 562
  5. 2024Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translationcited by 394
  6. 2016Multi-Agent Cooperation and the Emergence of (Natural) Languagecited by 185
  7. 2025Findings of the BabyLM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corporacited by 88
  8. 2020Multi-agent Communication meets Natural Language: Synergies between Functional and Structural Language Learningcited by 66
  9. 2018Emergent Communication through Negotiationcited by 62
  10. 2023Baby Llama: knowledge distillation from an ensemble of teachers trained on a small dataset with no performance penaltycited by 57
  11. 2024Findings of the Second BabyLM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corporacited by 48
  12. 2024GPT or BERT: why not both?cited by 27
14 more
  1. 2024Is Child-Directed Speech Effective Training Data for Language Models?cited by 24
  2. 2024Babysit A Language Model From Scratch: Interactive Language Learning by Trials and Demonstrationscited by 12
  3. 2024Searching for Structure: Investigating Emergent Communication with Large Language Modelscited by 10
  4. 2025Child-Directed Language Does Not Consistently Boost Syntax Learning in Language Modelscited by 9
  5. 2024NeLLCom-X: A Comprehensive Neural-Agent Framework to Simulate Language Learning and Group Communicationcited by 8
  6. 2024KidLM: Advancing Language Models for Children -- Early Insights and Future Directionscited by 8
  7. 2025Rubrik's Cube: Testing a New Rubric for Evaluating Explanations on the CUBE datasetcited by 8
  8. 2020Supervised Seeded Iterated Learning for Interactive Language Learningcited by 7
  9. 2024The Curious Case of Representational Alignment: Unravelling Visio-Linguistic Tasks in Emergent Communicationcited by 7
  10. 2025Investigating the Zone of Proximal Development of Language Models for In-Context Learningcited by 6
  11. 2025Simulating the Emergence of Differential Case Marking with Communicating Neural-Network Agentscited by 5
  12. 2021Towards Interactive Language Modelingcited by 4
  13. 2023Yes, this Way! Learning to Ground Referring Expressions into Actions with Intra-episodic Feedback from Supportive Teacherscited by 2
  14. 2025What is the Best Sequence Length for BABYLM?cited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 4; 2026: 5.

9 papers.

  1. 2026BabyLM Turns 4 and Goes Multilingual: Call for Papers for the 2026 BabyLM Workshopcited by 13
  2. 2025Language models as tools for investigating the distinction between possible and impossible natural languagescited by 0
  3. 2025Large language models have learned to use languagecited by 0
  4. 2025Large language models are not about natural languagecited by 0
  5. 2025Linguists should learn to love speech-based deep learning modelscited by 0
  6. 2026Are Language Models Models?cited by 0
  7. 2026Across the Levels of Analysis: Explaining Predictive Processing in Humans Requires More Than Machine-Estimated Probabilitiescited by 0
  8. 2026You Can't Fight in Here! This is BBS!cited by 0
  9. 2026Not-So-Strange Love: Language Models and Generative Linguistic Theories are More Compatible than They Appearcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Dialogue Is Not Enough to Make a Communicative BabyLM (But Neither Is Developmentally Inspired Reinforcement Learning)
  2. 2026Found in Conversation: LLMs Teach Themselves to Close the Multi-Turn Gap
  3. 2025EduDial: Constructing a Large-scale Multi-turn Teacher-Student Dialogue Corpus
  4. 2025Towards Developmentally Plausible Rewards: Communicative Success as a Learning Signal for Interactive Language Models
  5. 2025ConsistentChat: Building Skeleton-Guided Consistent Multi-Turn Dialogues for Large Language Models from Scratch
  6. 2024Talk Less, Interact Better: Evaluating In-context Conversational Adaptation in Multimodal LLMs
  7. 2025Beyond Single-Turn: A Survey on Multi-Turn Interactions with Large Language Models
  8. 2025A Survey on Multi-Turn Interaction Capabilities of Large Language Models