English

Post-training for Efficient Communication via Convention Formation

Computation and Language 2025-08-11 v1 Artificial Intelligence Machine Learning

Abstract

Humans communicate with increasing efficiency in multi-turn interactions, by adapting their language and forming ad-hoc conventions. In contrast, prior work shows that LLMs do not naturally show this behavior. We develop a post-training process to develop this ability through targeted fine-tuning on heuristically identified demonstrations of convention formation. We evaluate with two new benchmarks focused on this capability. First, we design a focused, cognitively-motivated interaction benchmark that consistently elicits strong convention formation trends in humans. Second, we create a new document-grounded reference completion task that reflects in-the-wild convention formation behavior. Our studies show significantly improved convention formation abilities in post-trained LLMs across the two evaluation methods.

Keywords

Cite

@article{arxiv.2508.06482,
  title  = {Post-training for Efficient Communication via Convention Formation},
  author = {Yilun Hua and Evan Wang and Yoav Artzi},
  journal= {arXiv preprint arXiv:2508.06482},
  year   = {2025}
}

Comments

Accepted to COLM 2025

R2 v1 2026-07-01T04:41:28.677Z