Data Overview

Lists, Posters, Stills, Loglines, Prompts, and Titles

The data were collected between October 2025 and March 2026. The image generation models used (and both available via ChatGPT's basic paid version's interface) were DALL-E 3 and ChatGPT 5's integrated image generator.

Foundational List

DATASET: 25 lists of 50 Lesbian Films generated by ChatGPT 5

  • Created by asking: "Make a list of 50 Lesbian Films" in 25 distinct, independent chats.
  • Columns 1-25 number the individual lists.

DATASET (related): A 4-tiered distribution of the 1256 film titles in the corpus of lists, organized into the following columns:

  • Film Title
  • Occurrences: Exact number of mentions in the dataset of 25 lists.
  • Group Name
    • Over 20 Occurrences (17 films). This list constitutes the basis for closer text and image database creation and analysis.
    • 10-19 Occurrences
    • 2–9 Occurrences
    • Single Occurrences
  • Group Number: The groups created above have been numbered 1-4

Posters, Stills, Loglines, Prompts, and Titles

(related to the 17 most-mentioned films)

Main Database (image and text)

The following data points were collected, created, and organized in a multimodal Main Database. Each column presents a DATASET excerpt-able for comparative analyses:

Assembled Available Data:

  • For each of the 17 film titles: IMDb link, Release Year, Country of Origin (Production), IMDb logline, IMDb poster, choice of one still from IMDb collection.

Generated Data: "Inspired by" the Top 17 film titles is divided into Sub-DATASETS. DATASETS 1-4 show results of engagement with DALL-E 3; DATASETS 5-11 with the ChatGPT-5 integrated image generator. Both models were available through ChatGPT's basic subscription in the fall of 2025.

  • DATASET 1 (image): DALL-E 3 — Create an image inspired by the Lesbian film [title].
  • DATASET 2 (text): DALL-E 3 — (grounded prompt for each image)
  • DATASET 3 (text): DALL-E 3 — Create a logline for a Lesbian film based on this first image in Dataset 1.
  • DATASET 4 (image): DALL-E 3 — Create a film poster for a Lesbian film with the logline: [add logline based on generated image]. Be sure to include a title.
  • DATASET 5 (image): GPT 5 — Create a still inspired by the Lesbian film [title].
  • DATASET 6 (text): GPT 5 — (to excerpt the grounded prompt:) Can you share how you adapted the prompt for this image?
  • DATASET 7 (text): GPT 5 — Create a logline for a Lesbian film based on the image [image in DATASET 5]?
  • DATASET 8 (image): GPT 5 — Create a film poster for a Lesbian film with the logline [add logline from DATASET 7]. Be sure to add a title.
  • DATASET 9 (text): new title created as part of poster in DATASET 8
  • DATASET 10 (image): GPT 5 — (word "Lesbian" omitted from request.) Create an image based on this: [add logline from DATASET 7]
  • DATASET 11 (image): GPT 5 — (word "Lesbian" omitted from request.) Create a film poster for: [title from DATASET 9]