Balancing Quality & Efficiency: LSA-DAT for Pseudocode-to-Code Conversion | CSCI 250 Paper Review

A presentation and review of the 2025 paper "Balancing Quality and Efficiency: An Improved Non-Autoregressive Model for Pseudocode-to-Code Conversion" by Xu, Zheng, Zou, Yu, and Gao (Journal of Systems & Software, Vol. 219).
In this video, we walk through the paper's core contributions: the LSA-DAT model, two data preprocessing strategies (Variable Replacement and Repetition Handling), and a new evaluation metric called Synthesis Potential (passp@N).

Видео Balancing Quality & Efficiency: LSA-DAT for Pseudocode-to-Code Conversion | CSCI 250 Paper Review канала Navraj Dhaliwal