​
Input

  • output (str): The generated text to be evaluated.
  • expectedOutput (str): The reference or ground truth text.

​
Output

  • Result (float): A distance score from 0 to infinity.

​
Interpretation

  • 0: The embeddings are identical.
  • Lower scores: The texts are more semantically similar.
  • Higher scores: The texts are more semantically different.
    The score is determined entirely by the one dimension where the embeddings differ the most.

​
Formula

Chebyshev(x,y)=max⁡i∣xi−yi∣\mathrm{Chebyshev}(x, y) = \max_i |x_i - y_i| Where i indexes the dimensions of the vectors.
This is a distance metric. Lower scores indicate greater similarity. It is particularly sensitive to the single largest deviation.

​
How It Works

The evaluator computes embeddings for both texts and then finds the maximum absolute difference across all corresponding dimensions of the vectors.

​
Use Cases

  • Identifying outliers in semantic similarity
  • Scenarios where the maximum deviation is more important than the average deviation
  • Logistics and chessboard-related distance calculations