Google

A Large Language Model that has been trained on extensive datasets of both source code and natural language can be effectively utilized for the task of code debugging.

Code Debugging with LLMs

You are working with the Python script below, which is intended to calculate the average value from a list of numbers. When you run the script with the input `[10, 20, 30, 40]`, you expect the output to be `25.0`, but instead, it prints `10.0`. Your task is to write a clear and effective prompt that you would submit to a large language model to help you identify and fix the error in your code.

Crafting an Effective Debugging Prompt

A developer is working on four different code snippets, each containing a bug. In which of the following scenarios would a large language model, trained on a vast corpus of code, be most effective at identifying and suggesting a fix for the error?

A developer has written the Python function below to find the smallest number in a list of positive integers. When they test it with the list `[10, 3, 8, 5]`, the function incorrectly returns `0` instead of the expected `3`.

```python
def find_smallest_number(numbers):
  smallest_so_far = 0
  for num in numbers:
    if num < smallest_so_far:
      smallest_so_far = num
  return smallest_so_far
```

They ask a large language model for help debugging it. Analyze the four potential responses from 

Learn Before

Related