binabase

// Bina is Comprehension (in Hebrew)

We read binabase comprehension files to solve logical problems and complex textual instructions.

Replacing LM NNs with rule based natural language processing.

binabase files

The binabase comprehension files are created with the assistance of LLMs and humans.

In future versions we will attempt to automate the process to several fields (some related to the pashute/aidash project):

1. General OCR to raw letter-component options, for pre-processing into results.

2. Specific field textbook OCR. (Using #1 and binabase comprehension files)

3. Specific field manuscript OCR