Large Vocabulary Size Improves Large Language Models

This paper empirically investigates the relationship between subword vocabulary size and the performance of large language models (LLMs) to provide insights on how to define the vocabulary size. Experimental results show that larger vocabulary sizes lead to better performance in LLMs. Moreover, we consider a continual training scenario where a pre-trained language model is trained on a different target language. We introduce a simple method to use a new vocabulary instead of the pre-defined one. We show that using the new vocabulary outperforms the model with the vocabulary used in pre-training.

Lower Gravity Demonstratable Testbed for Space Robot Experiments

In developing mobile robots for exploration on the planetary surface, it is crucial to evaluate the robot's performance, demonstrating the harsh environment in which the robot will actually be deployed. Repeatable experiments in a controlled testing environment that can reproduce various terrain and gravitational conditions are essential. This paper presents the development of a minimal and space-saving indoor testbed, which can simulate steep slopes, uneven terrain, and lower gravity, employing a three-dimensional target tracking mechanism (active xy and passive z) with a counterweight.

Decreasing the size of the Restricted Boltzmann machine

We propose a method to decrease the number of hidden units of the restricted Boltzmann machine while avoiding decrease of the performance measured by the Kullback-Leibler divergence. Then, we demonstrate our algorithm by using numerical simulations.