d1-2B decision model skims 1864 public-domain films to answer yes/no questions
vanstriendaniel · x · 2026-10-31
A demo by Daniel van Strien searches 1,864 public-domain Prelinger films by caption, then poses a yes/no question (e.g. "anyone wearing a mask?") to Liquid AI's d1-2B, a small decision model that checks 3 frames per candidate clip, 24 moments at a time, and jumps to the first clip it labels "yes". A screenshot shows a hit at 7:30 of a 1941 New Orleans Mardi Gras parade film with 0.67 confidence. Hosted on Hugging Face Spaces, requires sign-in and uses your own ZeroGPU quota.
More from Multimodal
- GVOID tool appears to run WAN 2.7 with new Cinematic mode and start-end frame interpolation — MillsLittlePistachio · 2026-10-31
- Prompt template: manga-style character sheet with multi-angle views and expressions — azed_ai · 2026-10-31
- Qwen Image 2.1 2-step distill LoRA hits v1.5, enabling realtime editing on a single GPU — linoy_tsaban · 2026-10-31
- Glif demos end-to-end brand content with hundreds of castable AI models — Puzzleheaded-King584 · 2026-10-31
- Claude Opus drives Blender MCP to build 3D models straight from 3DGS — janusch_patas · 2026-10-31
- FastSDCPU Halloween release adds Fibo 2.0 Edit Turbo image editing via OpenVINO on CPU — simpleuserhere · 2026-10-31