Arbor: Tree Search as a Cognition Layer for Autonomous Agents
arXiv:2606.12563 (cs.AI)
Submitted on 10 Jun 2026
Authors: Neha Prakriya, Chaojun Hou, Zheng Gong, Huasha Zhao, Xi Zhao, Mou Li, Zhenyu Gu, Emad Barsoum
Abstract
Arbor is a multi-agent framework that introduces structured tree search as a cognition layer for autonomous agents operating in large, stateful action spaces. Unlike prior autonomous optimization systems that work on isolated targets with stateless evaluation, Arbor maintains an explicit search tree of scored hypotheses. This tree serves as the shared working memory across agents, evolving with every measurement. Failures are treated as diagnostic signals that reshape subsequent exploration, and the tree expands as prior successes shift the bottleneck distribution.
We validate Arbor on full-stack LLM inference optimization—a domain where achieving peak performance has historically required coordinated effort from engineering teams across the application, framework, compiler, kernel, and hardware stack. Arbor pairs an Orchestrator agent, which drives optimization by delegating to Domain Specialists across the inference stack, with a Critic agent that safeguards stability through root-cause analysis, introspection, and measurement validation. This checks-and-balances architecture ensures that neither agent can unilaterally drive the system.
Agent capabilities are decomposed into hard skills (domain expertise) and soft skills (coordination protocols that determine how contributions compose), enabling fully autonomous multi-day campaigns. Arbor achieves up to 193% inference throughput-latency Pareto improvement over vendor-optimized baselines. In contrast, a single agent without the harness plateaus at +33% throughput improvement and crashes irrecoverably within hours. Arbor generalizes to multiple generations of hardware platforms, and run-to-run variance is within 2 percentage points, demonstrating that the method is hardware-agnostic and reproducible.
Subjects: Artificial Intelligence (cs.AI)
Cite as: arXiv:2606.12563 [cs.AI]
DOI: https://doi.org/10.48550/arXiv.2606.12563
via ArXiv AI