Early Continue

The model receives the prompt (and optional system message). The run uses scorer contains_any with the JSON configuration below. Pass/fail and partial credit are determined entirely by that scorer against the model output; no human grading.

Refactor for readability without behavior change. Return only Python code.

def positive(nums):
    out = []
    for n in nums:
        if n > 0:
            out.append(n)
    return out

{
  "expected_contains": [
    "if n <= 0",
    "continue",
    "append"
  ]
}