AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

TWIGMA: A dataset of AI-Generated Images with Metadata From Twitter

Yiqun T. Chen, James Zou

NeurIPS 2023 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Compare images less similar to ArtBench neighbors with images most similar to ArtBench neighborsTWIGMA images, adjusted for follower category · TWIGMA; ArtBench similarity quintileslikes: 19% more likesImproved
Compare images in ArtBench similarity quintile 2 with images most similar to ArtBench neighborsTWIGMA images, adjusted for follower category · TWIGMA; ArtBench similarity quintileslikes: 12% more likesImproved
Compare images in ArtBench similarity quintile 3 with images most similar to ArtBench neighborsTWIGMA images, adjusted for follower category · TWIGMA; ArtBench similarity quintileslikes: 0.8% fewer likesHurt
Compare images in ArtBench similarity quintile 4 with images most similar to ArtBench neighborsTWIGMA images, adjusted for follower category · TWIGMA; ArtBench similarity quintileslikes: 2% more likesImproved
Compare images less similar to LAION neighbors with images most similar to LAION neighborsTWIGMA images, adjusted for follower category · TWIGMA; LAION similarity quintileslikes: 10% (95% CI: [9%, 10%])Improved
Compare images in LAION similarity quintile 2 with images most similar to LAION neighborsTWIGMA images, adjusted for follower category · TWIGMA; LAION similarity quintileslikes: 2% more (95% CI: [2%, 3%])Improved
Condition image pairs on identical prompts rather than different promptsDiffusionDB · DiffusionDBpairwise image-embedding distance: roughly halfHurt
Use longer, more detailed promptsDiffusionDB image generation · DiffusionDBwithin-prompt output variation: Kendall's τ: -0.1, p-values: 0.01Hurt
Compare AI-generated image distributions with non-AI image distributionsTWIGMA and DiffusionDB · TWIGMA, DiffusionDB, LAION, and ArtBenchdistributional distinctivenessImproved
Compare image variability across datasetsTWIGMA · TWIGMA, LAION, ArtBench, and DiffusionDBimage variationHurt
Compare TWIGMA themes over timeTWIGMA · Twitter images from 2021 to 2023prevalence of artistically sophisticated themesImproved
Compare TWIGMA themes over timeTWIGMA · Twitter images from 2021 to 2023prevalence of natural scenesHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy