Mini Minds: Exploring Bebeshka and Zlata Baby Models
This work addresses the problem of building efficient language models from limited data, relevant for applications in human language acquisition and practical AI tasks, but it is incremental as it builds on existing small-scale modeling approaches.
The authors tackled the BabyLM competition's Strict-Small track by developing two small language models, Bebeshka and Zlata, which achieved comparable performance to baseline models despite being half the scale, and explored their use in moral judgment tasks.
In this paper, we describe the University of Lyon 2 submission to the Strict-Small track of the BabyLM competition. The shared task is created with an emphasis on small-scale language modelling from scratch on limited-size data and human language acquisition. Dataset released for the Strict-Small track has 10M words, which is comparable to children's vocabulary size. We approach the task with an architecture search, minimizing masked language modelling loss on the data of the shared task. Having found an optimal configuration, we introduce two small-size language models (LMs) that were submitted for evaluation, a 4-layer encoder with 8 attention heads and a 6-layer decoder model with 12 heads which we term Bebeshka and Zlata, respectively. Despite being half the scale of the baseline LMs, our proposed models achieve comparable performance. We further explore the applicability of small-scale language models in tasks involving moral judgment, aligning their predictions with human values. These findings highlight the potential of compact LMs in addressing practical language understanding tasks.