T-DIG releases first text-conditioned diversity dataset with 20k filtered prompts
simi_97k · x · 2026-10-28
The T-DIG team built their dataset by filtering 20k prompts from HPD V2 (a log of human-written Stable Diffusion prompts) and synthesizing axes of variation with Gemini 3.8 Flash, seeded from human-written axes. They also released a test set curated from real user interactions with the T-DIG tool — together the first dataset for text-conditioned diversity in image generation. The tool suggests variation axes for a model's i.i.d. image set, lets you set a diversity strength, and iteratively regenerate or edit sets.
Related event: CMU Open-Sources T-DIG: Text-Conditioned Diversity for Image Generation(7 posts)→
More from Multimodal
- Zero-AI-at-runtime design generator hits run 8, new aesthetics expected by run 10 — Afinetheorem · 2026-10-28
- Painting with EmbeddingGemma 2: ~30 images/sec on Apple Neural Engine after 12h A100 training — GlennCameronjr · 2026-10-28
- Krea2 BBOX models released: 40k-image fullrank finetune with multi-character LoRA support — Amazing_Painter_7692 · 2026-10-28
- One-line Tekken-style prompt in HeyGen Video yields ad-grade 2K fight footage — aziz4ai · 2026-10-28
- Stunning AI-generated concept art casts goddess Sekhmet walking a post-apocalyptic wasteland — DionysianAgent · 2026-10-28
- Custom node pack brings looping H3 video generation to Comfy Cloud — PurzBeats · 2026-10-28