Semantic Latent Alignment for Neural Communications

This repository investigates semantic alignment and latent-space interoperability between independently trained neural communication models. It provides tools, metrics, and experimental pipelines to evaluate and mitigate semantic mismatches in image reconstruction tasks.

Setup & Development

Clone the repository and set up the environment using uv:

git clone <REPO_URL>
cd <REPO_NAME>

uv sync
uv run pre-commit install
uv run pre-commit run --all-files
uv run pytest