​
Input

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

​
Output

  • Result (int): An integer distance score, from 0 to the dimension of the vectors.

​
Interpretation

  • 0: The vectors are identical.
  • Higher scores: The vectors have more differing positions.
    The result is a discrete integer, not a continuous value.

​
Formula

Hamming(x,y)=∑i1[xi≠yi]\mathrm{Hamming}(x, y) = \sum_i \mathbb{1}[x_i \ne y_i] The sum counts the number of positions i where the elements differ.
This is a distance metric. Lower scores indicate greater similarity. The score is an integer representing the count of differing dimensions.

​
How It Works

The evaluator computes embeddings for both texts, compares them dimension by dimension, and counts the number of differing positions. For non-binary vectors, this can involve binarization (e.g., thresholding).

​
Use Cases

  • Comparing binary hashes or fingerprints of data
  • Error detection in telecommunications
  • Genetics (comparing DNA sequences)
  • Evaluating binary or categorical embeddings