PACIFIC: a framework for generating benchmarks to check Precise Automatically Checked Instruction Following In Code

Itay Dreyfuss, Antonio Abu Nassar, Samuel Ackerman +5

Large Language Model (LLM)-based code assistants have emerged as a powerful application of generative AI, demonstrating impressive capabilities in code generation and comprehension…