GPT Humanizer Project - Revision history

Saxtonmd77: /* GPT Humanizer Overview */

2025-12-26T01:14:17Z

GPT Humanizer Overview

← Older revision		Revision as of 01:14, 26 December 2025
Line 1:		Line 1:
	= ~~GPT Humanizer~~ Overview =		= Overview =

	This is a software program designed to humanize GPT text. In early versions, the name is Matt's GPT Humanizer. Version 1.0 was developed in December 2025. It currently only restricts the generation of words that regularly appear in GPT output. The addition of features that further humanize generated text is already being explored.		This is a software program designed to humanize GPT text. In early versions, the name is Matt's GPT Humanizer. Version 1.0 was developed in December 2025. It currently only restricts the generation of words that regularly appear in GPT output. The addition of features that further humanize generated text is already being explored.

Saxtonmd77: /* Usage */

2025-12-26T01:13:50Z

Usage

← Older revision		Revision as of 01:13, 26 December 2025
Line 27:		Line 27:
	* Choose a model, or accept the default model, gpt-4o-mini		* Choose a model, or accept the default model, gpt-4o-mini
	* Set the temperature between 1 (most creative) and 0 (most deterministic)		* Set the temperature between 1 (most creative) and 0 (most deterministic)
	* Click the ''''Send'''' button		* Click the '''Send''' button
	* The text will generate in the last text area on the page		* The text will generate in the last text area on the page

Saxtonmd77: /* Installation */

2025-12-26T01:13:13Z

Installation

← Older revision		Revision as of 01:13, 26 December 2025
Line 17:		Line 17:
	== Installation ==		== Installation ==
	# Create and activate a virtual environment on your OS		# Create and activate a virtual environment on your OS
	## Install dependencies from '''requirements.txt'''		# Install dependencies from '''requirements.txt'''
	### Export '''OPENAI_API_KEY''' in the OS shell		# Export '''OPENAI_API_KEY''' in the OS shell
	~~###~~# Run the server with Uvicorn		# Run the server with Uvicorn

	== Usage ==		== Usage ==

Saxtonmd77: /* Installation */

2025-12-26T01:12:54Z

Installation

← Older revision		Revision as of 01:12, 26 December 2025
Line 16:		Line 16:

	== Installation ==		== Installation ==
	1. Create and activate a virtual environment on your OS		# Create and activate a virtual environment on your OS
	2. Install dependencies from '''requirements.txt'''		## Install dependencies from '''requirements.txt'''
	3. Export '''OPENAI_API_KEY''' in the OS shell		### Export '''OPENAI_API_KEY''' in the OS shell
	4. Run the server with Uvicorn		#### Run the server with Uvicorn

	== Usage ==		== Usage ==

Saxtonmd77 at 01:10, 26 December 2025

2025-12-26T01:10:25Z

← Older revision		Revision as of 01:10, 26 December 2025
Line 17:		Line 17:
	== Installation ==		== Installation ==
	1. Create and activate a virtual environment on your OS		1. Create and activate a virtual environment on your OS
	2. Install dependencies from ''''requirements.txt''''		2. Install dependencies from '''requirements.txt'''
	3. Export ''''OPENAI_API_KEY'''' in the OS shell		3. Export '''OPENAI_API_KEY''' in the OS shell
	4. Run the server with Uvicorn		4. Run the server with Uvicorn

Saxtonmd77 at 01:09, 26 December 2025

2025-12-26T01:09:50Z

← Older revision		Revision as of 01:09, 26 December 2025
Line 1:		Line 1:
	= GPT Humanizer Overview =		= GPT Humanizer Overview =

	This ~~project exists~~ to ~~explore human-centered control over LLM output without turning~~ the ~~system into a policy engine or a brittle ruleset.~~		This is a software program designed to humanize GPT text. In early versions, the name is Matt's GPT Humanizer. Version 1.0 was developed in December 2025. It currently only restricts the generation of words that regularly appear in GPT output. The addition of features that further humanize generated text is already being explored.

	~~The goal~~ is ~~not to censor language, but to:~~
	* Shape tone and wording deliberately
	* Preserve meaning while avoiding specific terms
	* Create space for “humanizing” transformations downstream

	'~~''Design priorities'''~~
	* Simplicity over abstraction
	* Observability over automation
	* Clear separation between generation~~, evaluation, and rewriting~~
	* Minimal UI that ~~reveals behavior rather than hides it~~

	~~This is a foundation, not a product~~.

	~~Future work may include additional rewrite strategies, tone shaping, and documentation-first development, but the system should remain understandable end-to-end.~~




	~~A lightweight FastAPI service~~ that ~~generates~~ text ~~via an LLM, applies banned-word detection, and optionally rewrites output to preserve meaning while avoiding specified terms. A minimal headless web UI~~ is ~~included for interactive use and inspection of requests and responses~~.

	== Features ==		== Features ==
	* Plain-text ~~generation~~ endpoint		* Minimal headless UI for dashboard-like controls
	* ~~Optional banned-word filtering with rewrite pass~~		* Plain-text output with the endpoint within the UI
	* ~~Temperature and token controls~~		* Standard list of restricted words
	* ~~Headless browser UI~~ for ~~prompt entry~~ and ~~testing~~		* Simple addition of more restricted words
	* ~~No client-side model~~ logic		* Controls for model selection, temperature, and token usage
			* Model logic is entirely on the server

	== Requirements ==		== Requirements ==
Line 34:		Line 15:
	* An OpenAI API key		* An OpenAI API key

	== ~~Setup~~ ==		== Installation ==
	1. Create and activate a virtual environment		1. Create and activate a virtual environment on your OS
	2. Install dependencies from `requirements.txt`		2. Install dependencies from ''''requirements.txt''''
	3. Export `OPENAI_API_KEY` in the shell		3. Export ''''OPENAI_API_KEY'''' in the OS shell
	4. Run the server with Uvicorn		4. Run the server with Uvicorn

	== Usage ==		== Usage ==
	* ~~API endpoint: `POST /generate`~~		* Access the UI: open http://127.0.0.1:8080 in a browser window
	* UI: open `http://127.0.0.1:8080/` in a browser		* Write a prompt in the first text area
	* ~~Health check: `GET /healthz`~~		* Insert additional restricted words in the next text area, one line at a time
			* Choose a model, or accept the default model, gpt-4o-mini
	The `/generate~~` endpoint returns plain~~ text ~~only.~~		* Set the temperature between 1 (most creative) and 0 (most deterministic)
			* Click the ''''Send'''' button
			* The text will generate in the last text area on the page

Saxtonmd77: Created page with "= GPT Humanizer Overview = This project exists to explore human-centered control over LLM output without turning the system into a policy engine or a brittle ruleset. The goal is not to censor language, but to: * Shape tone and wording deliberately * Preserve meaning while avoiding specific terms * Create space for “humanizing” transformations downstream '''Design priorities''' * Simplicity over abstraction * Observability over automation * Clear separation betwee..."

2025-12-26T00:29:09Z

Created page with "= GPT Humanizer Overview = This project exists to explore human-centered control over LLM output without turning the system into a policy engine or a brittle ruleset. The goal is not to censor language, but to: * Shape tone and wording deliberately * Preserve meaning while avoiding specific terms * Create space for “humanizing” transformations downstream '''Design priorities''' * Simplicity over abstraction * Observability over automation * Clear separation betwee..."

New page

= GPT Humanizer Overview =

This project exists to explore human-centered control over LLM output without turning the system into a policy engine or a brittle ruleset.

The goal is not to censor language, but to:
* Shape tone and wording deliberately
* Preserve meaning while avoiding specific terms
* Create space for “humanizing” transformations downstream

'''Design priorities'''
* Simplicity over abstraction
* Observability over automation
* Clear separation between generation, evaluation, and rewriting
* Minimal UI that reveals behavior rather than hides it

This is a foundation, not a product.

Future work may include additional rewrite strategies, tone shaping, and documentation-first development, but the system should remain understandable end-to-end.

A lightweight FastAPI service that generates text via an LLM, applies banned-word detection, and optionally rewrites output to preserve meaning while avoiding specified terms. A minimal headless web UI is included for interactive use and inspection of requests and responses.

== Features ==
* Plain-text generation endpoint
* Optional banned-word filtering with rewrite pass
* Temperature and token controls
* Headless browser UI for prompt entry and testing
* No client-side model logic

== Requirements ==
* Python 3.11+
* An OpenAI API key

== Setup ==
1. Create and activate a virtual environment
2. Install dependencies from `requirements.txt`
3. Export `OPENAI_API_KEY` in the shell
4. Run the server with Uvicorn

== Usage ==
* API endpoint: `POST /generate`
* UI: open `http://127.0.0.1:8080/` in a browser
* Health check: `GET /healthz`

The `/generate` endpoint returns plain text only.