arXiv cs.AI· Weidong Tang, Ruiheng Wang, Jingxuan Zhang, Hexin Wang, Haoyuan Chen, Zhengfeiyang Zhang, Yixing Ma, Zichen Jiang, Ziyu Guo, Haodong Li, Zhengbo Jiao, Zhiqing Cui, Xinxiang Yin, Shuo Liang, Zhenglin Wan, Hongzhan Lin, Wangbo Zhao Kaipeng Zhang, Pengfei Zhou, Odin Zhang, Yang You·· 2 小时前AI 评分49
The Trinity：面向多模态智能体的能力驱动递归自我改进框架
The Trinity:Capability-Driven Recursive Self-Improvement for Multimodal Agents
AI 导读
论文提出 The Trinity，一个耦合世界生成与 Harness 演化的递归自我改进框架，由 Creator、Explorer 和 Evolver 三个角色组成，将任务树重组为层次化能力树，并通过 Capability-Driven Real2Sim2Real 构建可验证的模拟世界进行能力学习。
正文 · 原文
Authors:Weidong Tang, Ruiheng Wang, Jingxuan Zhang, Hexin Wang, Haoyuan Chen, Zhengfeiyang Zhang, Yixing Ma, Zichen Jiang, Ziyu Guo, Haodong Li, Zhengbo Jiao, Zhiqing Cui, Xinxiang Yin, Shuo Liang, Zhenglin Wan, Hongzhan Lin, Wangbo Zhao Kaipeng Zhang, Pengfei Zhou, Odin Zhang, Yang You
Abstract:Multimodal intelligence can be viewed as human-guided co-evolution between worlds and agents through interaction. However, scaling this process faces two bottlenecks: constructing heterogeneous, verifiable environments is costly, and manual Harness iteration cannot keep pace with growing task diversity and complexity. A key insight is that tasks with different modalities and surface forms can share the same underlying capability requirements. We therefore introduce The Trinity, a recursive self-improvement framework coupling world generation with Harness evolution through three roles: Creator, Explorer, and Evolver. Drawing on verified execution trajectories, Evolver shifts the basis of task organization from surface categories to shared requirements, reorganizing the original task tree into a hierarchical capability tree. We introduce Capability-Driven Real2Sim2Real: for heterogeneous tasks at the same node, Creator abstracts their common requirements into simplified, verifiable simulated worlds for capability learning, while evaluation on original tasks measures transfer. Along this capability tree, Explorer progressively evolves reusable capability modules through selective inheritance, forming a Harness tree whose validated updates accumulate across iterations within a fixed topology. The resulting Harnesses specialize in different capabilities, motivating a Capability-Aware Harness Router that uses public task information to select compatible module configurations from the tree. Our evolved Harnesses improve mean performance by 34 points across 28 training benchmarks and 26 points across 18 unseen benchmarks, including two real-world embodied tasks.
| Subjects: | Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2610.36077 [cs.AI] |
| (or arXiv:2610.36077v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.36077 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Weidong Tang [view email]
[v1]
Mon, 26 Oct 2026 18:03:34 UTC (42,027 KB)
来源：arXiv cs.AI · arxiv.org