الملخص
In recent years, LLM-based multi-agent systems have been widely applied to orchestrate tool-using agents into executable communication graphs. However, existing self-evolving orchestration still faces key challenges, including post-hoc evolution that revises the team only after the trajectory ends, credit diffusion that gives every action the same terminal advantage under confounded baselines, and skill admission that is uncalibrated and never retired. To address these challenges, we propose EvoSteer, a new paradigm of Online Self-Evolving Graph Orchestration -- the orchestrator builds a running team and repairs its plausible but failing steps from execution features and a learned value estimate. To support this paradigm, we introduce Anchored Trajectory Balance (AnchorTB), a regression-style flow-matching loss that assigns each orchestration action a coefficient by balancing subtrajectories against a frozen reference. Built on the learned flow, we further propose Validated Skill Admission, in which a candidate skill is tried before promotion and promoted only if paired evidence passes a sequential test under a shared nominal testing budget. Moreover, AnchorTB combines measured task-level reference reward statistics with prefix-dependent corrections. Experimental results on twelve datasets show that EvoSteer significantly outperforms baselines across question answering, mathematical reasoning, code generation, and interactive decision making. Our code is available at https://github.com/beita6969/evosteer.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Zhang, M., Zhang, H., Huang, Q., Wang, Z., Guo, P., Zhang, Y., Zhu, J., & Tang, X. (2026). EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignment. https://omanscience.com/ar/articles/evosteer-online-self-evolving-graph-orchestration-via-reference-anchored-credit-assignment
MLA 9
Zhang, Mingda, et al. "EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignment." https://omanscience.com/ar/articles/evosteer-online-self-evolving-graph-orchestration-via-reference-anchored-credit-assignment.
شيكاغو (المؤلف–التاريخ)
Zhang, Mingda, Hanwen Zhang, Qiang Huang, Zijia Wang, Pengfei Guo, Yuchen Zhang, Jionghao Zhu, and Xiaoying Tang. 2026. "EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignment." https://omanscience.com/ar/articles/evosteer-online-self-evolving-graph-orchestration-via-reference-anchored-credit-assignment.
هارفارد
Zhang, M., Zhang, H., Huang, Q., Wang, Z., Guo, P., Zhang, Y., Zhu, J. and Tang, X. (2026) 'EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignment', Available at: https://omanscience.com/ar/articles/evosteer-online-self-evolving-graph-orchestration-via-reference-anchored-credit-assignment.
فانكوفر
Zhang M, Zhang H, Huang Q, Wang Z, Guo P, Zhang Y, et al. EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignment. https://omanscience.com/ar/articles/evosteer-online-self-evolving-graph-orchestration-via-reference-anchored-credit-assignment
IEEE
M. Zhang, H. Zhang, Q. Huang, Z. Wang, P. Guo, Y. Zhang, J. Zhu, and X. Tang, "EvoSteer: Online Self-Evolving Graph Orchestration via Reference-Anchored Credit Assignment," https://omanscience.com/ar/articles/evosteer-online-self-evolving-graph-orchestration-via-reference-anchored-credit-assignment.