More scripts by Transys
Free · Python · Runs locally
LLM Explainer
Look inside a real large language model (GPT-2). See how it splits text into tokens, predicts the next word, writes one token at a time, uses attention, and stores meaning as vectors. Six interactive tabs in your browser.
- Install once:
pip install torch transformers gradio matplotlib - Run:
python llm_explainer.py - Your browser opens the app. The model (~350 MB) downloads on first run.