Deep Search with Hierarchical Meta-Cognitive Monitoring Inspired by Cognitive Neuroscience

TL;DR

The paper proposes DS-MCM, a hierarchical metacognitive framework for deep-search agents that couples a Fast Consistency Monitor with a Slow Experience-Driven Monitor to regulate reasoning and retrieval under uncertainty. By grounding fast monitoring in the relationship between external evidence uncertainty and internal reasoning uncertainty, and by leveraging an experience mémoire to guide reflective interventions, DS-MCM achieves notable improvements across diverse backbones and benchmarks, often surpassing proprietary systems. Empirical results show consistent gains, modest overhead, and strong generalization of the memory-based interventions, underscoring the practical value of embedded metacognitive control in long-horizon information-seeking tasks.

Abstract

Deep search agents powered by large language models have demonstrated strong capabilities in multi-step retrieval, reasoning, and long-horizon task execution. However, their practical failures often stem from the lack of mechanisms to monitor and regulate reasoning and retrieval states as tasks evolve under uncertainty. Insights from cognitive neuroscience suggest that human metacognition is hierarchically organized, integrating fast anomaly detection with selectively triggered, experience-driven reflection. In this work, we propose Deep Search with Meta-Cognitive Monitoring (DS-MCM), a deep search framework augmented with an explicit hierarchical metacognitive monitoring mechanism. DS-MCM integrates a Fast Consistency Monitor, which performs lightweight checks on the alignment between external evidence and internal reasoning confidence, and a Slow Experience-Driven Monitor, which is selectively activated to guide corrective intervention based on experience memory from historical agent trajectories. By embedding monitoring directly into the reasoning-retrieval loop, DS-MCM determines both when intervention is warranted and how corrective actions should be informed by prior experience. Experiments across multiple deep search benchmarks and backbone models demonstrate that DS-MCM consistently improves performance and robustness.

Figures (3)

• Figure 1: Cognitive neuroscience-inspired hierarchical metacognitive monitoring for deep search agents. Top: Humans rely on a fast, implicit monitor to detect anomalies during execution, which selectively triggers a slow, experience-driven monitor for reflective correction. Middle: Standard deep search agents lack such monitoring, allowing early cognitive errors to propagate across reasoning steps. Bottom: Our DS-MCM integrates the fast consistency monitor and slow experience-driven monitor into the deep search loop, enabling timely detection and correction of unreliable execution (Illustrative examples sampled from BrowseComp-Plus chenbrowsecomp).
• Figure 2: Overview of Deep Search with Meta-Cognitive Monitoring (DS-MCM). DS-MCM augments a standard ReAct-based deep search agent with an explicit metacognitive monitoring module. At each step, a Fast Consistency Monitor (§\ref{['sec:fast_monitor']}) checks whether reasoning uncertainty is calibrated with the uncertainty of retrieved evidence. When abnormal mismatches are detected, a Slow Experience-Driven Monitor (§\ref{['sec:slow_monitor']}) is triggered to perform reflective diagnosis using success and failure experiences stored in metacognitive memory, and to issue corrective guidance for subsequent steps. The memory is continuously updated and consolidated online, enabling efficient and experience-driven regulation of deep search execution.
• Figure 3: Sensitivity analysis of DS-MCM with respect to key hyperparameters.