Skip to content

Exploring LLMs' Abiliity To Generate & Use Personal Contextual Data

View on GitHub

alt text

One of the most interesting use-cases for LLMs is their ability to leverage personal information in order to provide actionable guidance:

For example:

  • Provided with a resume, an LLM can suggest opportunities for career advancement, upskilling, etc
  • Provided with your list of hobbies, an LLM can suggest opportunities to pursue those in a new city
  • Provided with your purchasing wishlist from Amazon, an LLM could suggest how you can optimise your budget to buy things you'll enjoy the most

As part of my wider interest in exploring LLMs (and open-sourcing the journey) I thought it would be interesting to share some of these. However, for obvious reasons, I can't freely disclose my own personal data.

The solution that came to mind:

Using an LLM to generate an alias and using that alias to show how LLMs can do all of these things (and more).

"David Rosen" is loosely based around my own biography with minor details changed and some invented. All information related to this identity is fictitious.

Prompts and outputs documented here are shared purely with the intention that they might help others to explore how LLMs could be used for good when guided by small chunks of personal context.

Index Of Demos

Repo Layout

The main part of the repo is demos which is where I log some prompts and outputs to try to highlight some of the potential use-cases.

My focus is on synthesising well-written prompts (a work in progress!) and specific contextual data. I find that this formula can produce very interesting results that often far exceed, in utility, those which would be achieved without the added context.

I try to be as thorough as possible when documenting my various experiments with large language models (LLMs).

For that reason, I've added, to this repo, the prompts I used to construct the synthetic personal identity of 'David Rosen'. These are gathered in development to make them easy to distinguish from the prompts (and outputs) that demonstrate potential use-cases.

Author

LLM Experimenteur: Daniel Rosehill
(public at danielrosehill dot com)

Licensing

This repository is licensed under CC-BY-4.0 (Attribution 4.0 International) License

Summary of the License

The Creative Commons Attribution 4.0 International (CC BY 4.0) license allows others to: - Share: Copy and redistribute the material in any medium or format. - Adapt: Remix, transform, and build upon the material for any purpose, even commercially.

The licensor cannot revoke these freedoms as long as you follow the license terms.

License Terms

  • Attribution: You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
  • No additional restrictions: You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.

For the full legal code, please visit the Creative Commons website.