H3-World turns MiniMax H3 into a controllable world simulator with 8k gameplay samples
sachasayan · reddit · 2026-09-02
H3-World repurposes MiniMax H3's pretrained text pathway into language-native world control: character and camera actions are composed as textual instructions and injected per video latent interval for temporally grounded control.
Remarkably efficient — just 8,000 gameplay samples, 10,000 LoRA steps, and 0.199% trainable parameters yield controllable character/camera motion that generalizes to unseen action compositions and visual scenarios. Paper (arXiv 2609.01560), code, model and project page are all public.
Related event: MiniMax Unveils H3-World, Its First Interactive World Model(6 posts)→
More from Multimodal
- MOSS-Image-SFT at 4 Steps Draws Hype: 320x320 to 2K Without Artifacts — Powerful_Evening5495 · 2026-10-18
- MOSS SFT vs Flux 3 Klein vs SD 3.0: Same-Prompt Comparison Exposes SD3's Age — ZootAllures9111 · 2026-10-18
- Dev builds Mumbai-inspired 3D manga background generator from OpenStreetMap data — CurieuxExplorer · 2026-10-18
- Omni 1'1 flash video demo generates atmospheric foggy Victorian haunted house — Dense-Instruction210 · 2026-10-18
- Flux 3 klein distilled tested: int8 workflow delivers surprisingly good 2MP images — Ok_Tale7582 · 2026-10-18
- AI-generated clip with cinematic truck-driver shot fuels 'feature film soon' talk — 1ili0 · 2026-10-18