This is a Code Competition
Submissions to this competition must be made through Notebooks. In order for the "Submit" button to be active after a commit, the following conditions must be met:

CPU Notebook <= 9 hours run-time
GPU Notebook <= 9 hours run-time
Internet access disabled
Freely & publicly available external data is allowed, including pre-trained models
Submission file must be named submission.csv
Submission runtimes have been slightly obfuscated. If you repeat the exact same submission you will see up to 15 minutes of variance in the time before you receive your score.

Your submissions will be evaluated on multiple criteria:
1. **Context Precision** (20%): How relevant is the retrieved context to the question?
2. **Answer Faithfulness** (20%): How well does the answer reflect the retrieved context?
3. **Answer Correctness** (40%): How accurate is the answer compared to the ground truth?
4. **Reference Accuracy** (20%): How precisely do the provided references match the relevant content?

### Submission Format
For each `ID` in the test set, you must provide a complete response. The submission file should be a CSV with a header and have the following format:
```csv
ID,context,answer,references
```

**Field Descriptions**:
- `ID`: The unique identifier for each question (note: we use 'ID' not 'query_id')
- `context`: The retrieved context from the textbook used to generate the answer
- `answer`: The answer generated by your system
- `references`: A JSON-formatted string containing sections and pages, formatted as:
  ```json
  {"sections": ["section1", "section2"], "pages": ["page1", "page2"]}
  ```

Example submission row:
```csv
1,"<retrieved context from book>","The scientific method in psychology involves...",{"sections": ["psychological_research/approaches_to_research"], "pages": ["41"]}
```