proomt

Search

Search posts, papers, and topics

All posts

Hugging Face Daily PapersMingyang Chen, Shengdong Chen, Xiaoxiao Fu1 min readpaperadvanced

Zing-0.5: Toward Playable Worlds with Real-Time Joint Action and Text Control

Read on arxiv.org

Summary

Zing‑0.5 is a 5 B autoregressive world model that lets users control generated environments in real time using both keyboard actions and text prompts. The paper introduces unified action‑text conditioning, segment‑level teacher distillation, and a low‑cost streaming inference pipeline that runs at 24 FPS (832×480) for about $0.009 per minute, achieving 81 % overall and 88.5 % consistency on a nav…

  • Unified conditioning merges magnitude‑aware keyboard inputs with temporally aligned text instructions, enabling joint control in a single sequence.
  • Event‑scale supervision uses a segment‑level teacher to distill knowledge into a block‑level causal student, improving incremental generation fidelity.
  • Four‑step generation with context‑preserving streaming runs at 24 FPS (832×480) for ~0.009 USD per stream‑minute, enabling low‑cost real‑time interaction.
  • On 158 WBench Navigation cases Zing‑0.5 scores 81.0 overall and 88.5 consistency, demonstrating reliable navigation and event control.

Game AI and interactive generative model engineers should care because it demonstrates a practical, low‑cost approach to real‑time, jointly controllable world generation.

7/10
machine learningreinforcement learninggenerative modelsworld modelsreal time

Related reading

  1. PhysStream: Streaming Physics-Grounded Video Generation with Structured Scene Memory and Fine-Grained Motion Control

    PhysStream introduces a two‑stage autoregressive video generator that uses online‑derived positional and tracking maps (structured scene memory) and sparse velocity‑increment signals to enable fine‑grained, physics‑grounded control of multi‑object tabletop scenes. It cuts motion distribution error by 33 % and trajectory error by 12 % versus strong baselines, and wins 85 % of human preference test…

    Hugging Face Daily Papersarxiv.org1 minpaper

  2. Building a Reliable Script-to-Video Pipeline for Modern Web Applications

    A practical guide to building a reliable, observable script‑to‑video pipeline. It advocates a staged architecture (scene manifest → asset generation → voice → rendering → encoding), validates AI‑generated JSON with Zod, persists project and scene state, uses explicit workflow states, runs heavy work in background queues, limits concurrency, and stores voice, captions, and assets separately to ena…

    SitePointsitepoint.com11 min

  3. Dynamically Scaled Activation Steering

    Dynamically Scaled Activation Steering (DSAS) is a method‑agnostic framework that learns per‑token, per‑layer scaling factors to turn existing activation‑steering interventions on only when a model is likely to produce undesired output (e.g., toxic text). The scaling can be optimized jointly with any steering function, improves the toxicity‑utility trade‑off on language models, transfers to text‑…

    Apple Machine Learning Researchapple.com1 minpaper

  4. MintAct: A Unified Visual Agent for Digital Environments

    MintAct is a family of vision-language models (2B-8B) that unifies UI grounding, multi-step navigation, and visual tool use across mobile, desktop, and web environments. It achieves state-of-the-art performance by leveraging a scalable environment and an asynchronous reinforcement learning infrastructure.

    Hugging Face Daily Papersarxiv.org1 minpaper

  5. RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents

    The paper presents RecreationWorld, a five‑platform framework that lets hybrid computer‑use agents learn by recreating the behavior of a running reference, and introduces RecreationBench, a 250‑task benchmark with programmatic and visual assertions. Experiments show GPT‑6 Astra reaches 58.1% overall but struggles with deeper programmatic tests, highlighting gaps in current agents.

    Hugging Face Daily Papersarxiv.org2 minpaper

  6. RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning

    RetireOPD introduces a self‑retiring on‑policy distillation framework for multi‑turn RL agents. A skill‑conditioned teacher is first trained with environment rewards, then a skill‑free student learns jointly via RL and token‑level distillation. The student automatically drops the teacher once its performance gap stops shrinking and it reaches a target success‑rate fraction, after which training c…

    Hugging Face Daily Papersarxiv.org1 minpaper