Skip to content
@AllSpark-Research

AllSpark Research

AllSpark at Xiaohongshu builds a unified, multimodal Agent Foundation Model for general-purpose capabilities and real-world applications.
README.md

Welcome to AllSpark 👋

AllSpark is RedNote(小红书)'s AI research team advancing the frontier of agent intelligence through unified, general-purpose multimodal Agent Foundation Models.

We explore how agents learn, reason, and improve through mid-training, supervised fine-tuning, reinforcement learning, and on-policy distillation.

We bring these capabilities to products across Xiaohongshu, helping users discover, create, and connect every day.

𝕏 X.com · 🤗 Hugging Face · 🧠 Models · 📄 Papers · 🔴 RedNote(小红书)

To Join Us

We recruit Research Interns year-round. If you are interested in joining AllSpark, please send your CV to allspark@xiaohongshu.com or apply through the Xiaohongshu recruitment portal.

Pinned Loading

  1. Iris Iris Public

    Climbing to the Search Frontier.

    Python 101 2

  2. AgentEnv AgentEnv Public

    Allspark Agent Data & Environment Synthesis

    Python 22

  3. Vela Vela Public

    Vela: User Behavior as Supervision Sparks LLM Intelligence — official paper, task templates, and prompts from AllSpark at Xiaohongshu.

    24 1

  4. AdaptEvo AdaptEvo Public

    Adaptive Agent Learning with Evolving Supervision

    Python 146 12

  5. D-3-MOPD D-3-MOPD Public

    D^3-MOPD: DYNAMIC DOMAIN SCHEDULING FOR EFFICIENT MULTI-TEACHER DISTILLATION

    Python 51 1

Repositories

Showing 8 of 8 repositories

People

This organization has no public members. You must be a member to see who’s a part of this organization.

Top languages

Loading…

Most used topics

Loading…