Oriel — AI & technology worth noticing, every day.

Hy-Nomas: Towards Free-form Interactive World Generation

Paper · https://arxiv.org/abs/2610.36729

Interactive world generation requires models to follow successive commands while preserving relevant history and action consequences. We present Hy-Nomas, a video world model with an 8B diffusion backbone that supports free-form text instructions and continuous camera control across multiple interaction turns. It generates variable-length segments conditioned on current inputs and cached history. Our central design aligns structured video captions with action-aware causal attention: descriptive context remains accessible across segments, while action instructions are restricted to their intended temporal windows. A vision-language Prompt Enhancer (PE) translates user requests and multimodal interaction histories into structured conditions for the next segment. A multi-stage curriculum progressively introduces causal generation, text interaction, camera control, and training on model-generated histories, followed by diffusion distillation. In human evaluation, Hy-Nomas shows superior performance over existing world models. We release model weights, code and benchmarks, together with a lightweight adaptation workflow, to support reproducible research on interactive world generation.

First seen by Oriel on .

AI summary

Hy-Nomas is an 8B diffusion-based video world model for generating interactive worlds from free-form text instructions and continuous camera control across multiple turns.

  • It generates variable-length video segments using current inputs and cached interaction history.
  • Structured video captions and action-aware causal attention preserve descriptive context while limiting actions to their intended temporal windows.
  • A vision-language Prompt Enhancer converts user requests and multimodal interaction histories into structured conditions for subsequent segments.
  • Training progressively adds causal generation, text interaction, camera control, model-generated histories, and diffusion distillation.
  • Human evaluation found superior performance over existing world models, and the authors release weights, code, benchmarks, and a lightweight adaptation workflow.

#world models #video generation #diffusion models #multimodal interaction

Generated by gpt-5.6-sol. Check the original for details.

Metric trends

Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained

Points

Score reported by the source.

  • Hugging Face Papers: 112, latest observation ; +10 over observed period since

Comments

Comment count reported by the source.

  • Hugging Face Papers: 2, latest observation ; +1 over observed period since

Aliases

SourceExternal IDURL
Hugging Face Papers2610.36729https://arxiv.org/abs/2610.36729 Read discussion