This repository extends Google Research's Sufficient Input Subsets (SIS) implementation with research-engineering tools for model-agnostic explainability. SIS can explain neural networks through a ...
Standalone scripts for clinical data extraction, de-identification, data cleaning/imputation, text embedding (LLM-based, BGE-M3-based, and Doc2Vec-based), and multi-algorithm classification modeling ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results