1 paper · 1 filter
Yixuan Li, Julian Parsert, Elizabeth Polgreen
Pre-trained Large Language Models (LLMs) are beginning to dominate the discourse around automatic code generation with natural language specifications. In contrast, the best-perfor…