Find our extension code (LLM + chrome extension) code on our extension branch and the recursive summarization code in main.
Content was scraped from WCAG guidelines (linked here) and converted into json objects & summarized as a tree. This meant summarizing the end-nodes and using that summary to work our way up.
Our prompt for OpenAI's GPT in this case was:
SUMMARIZE_PROMPT = """Write a {paragraphs} paragraph the following about accessibility guidelines into readable concise clear prose without any special character, focusing on what a developer would need to specifically implement while ignoring section titles and numbers:
Text: {text}
Summary of {paragraphs} paragraphs:"""
SUMMARIZE_PROMPT_TOKENS = len(tiktoken.get_encoding("cl100k_base").encode(SUMMARIZE_PROMPT))Tiktoken embeddings was used in case the input was too large for the model (davinci-003) over 2097 tokens. The website was generated through HTML where clicks expanded down each "tree" branch.