Write a Python function `frequency_lists` that takes a list of lists as input and returns a dictionary where the keys are the unique elements from the flattened list, and the values are the counts of their occurrences.
#### Example Usage
```python [main.nopy]
frequency_lists([[1, 2, 3, 2], [4, 5, 6, 2], [7, 8, 9, 5]])
# Output: {1: 1, 2: 3, 3: 1, 4: 1, 5: 2, 6: 1, 7: 1, 8: 1, 9: 1}
frequency_lists([[1, 2], [2, 3], [3, 4]])
# Output: {1: 1, 2: 2, 3: 2, 4: 1}
```
#### Constraints
- The input will always be a list of lists.
- The inner lists can contain integers.
- The function should handle empty lists gracefully.