自进化多模态智能体的跨场景综述与分析
A Cross-Scenario Survey and Analysis of Self-Evolving Multimodal Agents
AI 导读
一篇 arXiv 综述（arXiv:2611.14256，44 页、15 张表、4 幅图）研究多模态智能体自我进化机制在四大场景间的迁移不对称性：多模态推理、多模态生成、GUI 与 Web 导航、具身操作。作者以“经验生成—反馈获取—更新巩固”三阶段循环为框架，将各机制判定为可迁移、有条件、受阻或冗余，并把不对称归因于相反迁移方向上稀缺因素与任务结构资源的组合差异。
正文 · 原文
Abstract:Recent work shows that multimodal agents increasingly self-improve without recurring human annotation. However, each self-improvement mechanism is designed and evaluated within a single scenario, leaving its preconditions unverified under cross-scenario transfer. This survey investigates the structural source of asymmetric mechanism transfer across four scenarios: multimodal reasoning, multimodal generation, GUI and Web navigation, and embodied manipulation. We first formalize the self-evolving multimodal agent by four criteria and a differentia of self-reference, and adopt the three-stage self-evolving loop of experience generation, feedback acquisition, and update consolidation as our analytical framework to position the scenarios along ceilings on rollout count and feedback fidelity. Comparing self-evolving with fixed loops within each scenario separates the limitations that the scenario imposes from those inherent in the self-evolving design. We then recast each mechanism as a trade-off that raises a scarce factor by presupposing a task-structural resource, such as output equivalence or checkable outcomes, and judge it portable, conditional, blocked, or redundant across scenarios. We attribute the asymmetry to divergent combinations of scarce factors and resources across opposite transfer directions, and identify the matched-budget measurements that could overturn these still untested verdicts.
| Comments: | 44 pages, 15 tables, 4 figures |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.14256 [cs.AI] |
| (or arXiv:2611.14256v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.14256 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jingjing Jiang [view email]
[v1]
Tue, 10 Nov 2026 16:04:02 UTC (205 KB)
来源：arXiv cs.AI · arxiv.org