{
  "context": "Tests whether the response correctly identifies common Python memory leak patterns in a pandas/web service context.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Identifies DataFrame retention as likely cause",
      "description": "Points out that DataFrames may be held in memory via references (global variables, caches, closures, or uncleared lists)",
      "max_score": 15
    },
    {
      "name": "Recommends diagnostic tools",
      "description": "Suggests tracemalloc, objgraph, memory_profiler, or similar tools to diagnose the leak",
      "max_score": 12
    },
    {
      "name": "Mentions garbage collection considerations",
      "description": "Discusses gc module, reference cycles, or del to explicitly free DataFrames",
      "max_score": 10
    },
    {
      "name": "Suggests chunked processing",
      "description": "Recommends processing large CSVs in chunks (pd.read_csv chunksize) instead of loading all at once",
      "max_score": 10
    },
    {
      "name": "No incorrect information",
      "description": "Response does not contain factually wrong statements about Python memory management or pandas",
      "max_score": 12
    }
  ]
}