2506.07807v2

# A Proposal to Extend the Common Model of Cognition with Metacognition **Authors**: John Laird, Christian Lebiere, Paul Rosenbloom, Andrea Stocco institutetext: Contact: John.Laird@cic.iqmri.org ## Abstract The Common Model of Cognition (CMC) provides an abstract characterization of the structure and processing required by a cognitive architecture for human-like minds. We propose a unified approach to integrating metacognition within the CMC. We propose that metacognition involves reasoning over explicit representations of an agent’s cognitive capabilities and processes in working memory. Our proposal exploits the existing cognitive capabilities of the CMC, making minimal extensions in the structure and information available within working memory. We provide examples of metacognition within our proposal. Keywords: Common Model of Cognition, Cognitive Architecture, Metacognition ## 1 Introduction The Common Model of Cognition (CMC) [9] was developed as an abstract consensus model of human-like minds, derived from the computational structures and representations of cognitive architectures [6], and informed by our knowledge of the human mind and brain. The CMC specifies the fixed architectural structures for encoding, maintaining, using, and acquiring knowledge to produce behavior, emphasizing routine cognition. Here, we propose extending the CMC to include metacognition [3, 4, 5, 7, 12, 17]. Broadly speaking, metacognition encompasses reasoning about any aspect of cognition, including reasoning, memory, perception [13], motor skills, and learning. It can include partial or even incorrect theories of cognition, such as reasoning about perceived ESP capabilities. We focus on metacognition related to an agent’s own cognition. We also discuss metareasoning: reasoning about reasoning as a restricted form of metacognition. Examples of metacognition include: introspective monitoring, deliberate decision making, deliberate learning [10], predictive and hypothetical reasoning, retrospective reasoning, strategy selection, and self-explanation. Figure 1 shows two approaches to metacognition in cognitive architectures. In the hierarchical approach, also referred to as the Nelson and Narens model [11], specialized modules are added “on top” of cognition to implement metacognition, as exemplified by MIDCA [2] and Clarion [16]. Those modules monitor cognition, reason about it, and modify its processing. They access the current state of cognitive processing, the histories of processing, and representations of the agent’s procedural knowledge that drives cognition, with reasoning and metareasoning operating in parallel without intermixing. <details> <summary>extracted/6529175/Figures/MCArch1.png Details</summary>  ### Visual Description ## Diagram: Comparison of Hierarchical vs. Unified Architectures ### Overview The image displays two architectural diagrams side-by-side, labeled "Hierarchical" and "Unified." These diagrams illustrate the structural organization of cognitive or AI system components—specifically how knowledge, reasoning, and situational awareness interact with an environment. The diagrams contrast a multi-layered, modular approach (Hierarchical) with a consolidated, integrated approach (Unified). ### Components/Axes **Left Diagram: Hierarchical** * **Top Box (Dark Blue):** "Metareasoning Knowledge" * **Middle Box (Medium Blue):** "Reasoning Knowledge" * **Bottom Box (Grey):** "Situation Representation" * **Container:** A light-grey rounded rectangle encloses both the "Reasoning Knowledge" and "Situation Representation" boxes. * **Base (Black Oval):** "Environment" * **Flow Indicators:** * Circular arrow between "Metareasoning Knowledge" and "Reasoning Knowledge." * Circular arrow between "Reasoning Knowledge" and "Situation Representation." * Vertical arrows connecting "Environment" to the "Situation Representation" box (Upward arrow for input, Downward arrow for output). **Right Diagram: Unified** * **Top Box (Medium Blue):** "Reasoning and Metareasoning Knowledge" * **Bottom Box (Grey):** "Situation and Partial Reasoning State Representation" * **Base (Black Oval):** "Environment" * **Flow Indicators:** * Circular arrow between the Top Box and Bottom Box. * Vertical arrows connecting "Environment" to the bottom box (Upward arrow for input, Downward arrow for output). ### Detailed Analysis **Hierarchical Model (Left):** * **Structure:** This model is stratified into three distinct functional layers. * **Flow:** The system implies a top-down or supervisory relationship. The "Metareasoning Knowledge" acts as a distinct layer above the "Reasoning Knowledge." The grouping of the lower two layers suggests that "Reasoning" and "Situation Representation" operate within a specific operational boundary, distinct from the "Metareasoning" layer. * **Interaction:** The circular arrows indicate bidirectional feedback loops between adjacent layers. The environment interacts exclusively with the lowest layer ("Situation Representation"). **Unified Model (Right):** * **Structure:** This model is condensed into two functional layers. * **Flow:** The system implies an integrated approach. "Reasoning" and "Metareasoning" are combined into a single block, suggesting that the distinction between the two is either removed or handled internally within that block. * **Interaction:** The "Situation Representation" is expanded to include "Partial Reasoning State," suggesting that the representation layer now carries more information (the state of the reasoning process itself) than in the Hierarchical model. The environment interacts with this combined representation layer. ### Key Observations * **Consolidation:** The Unified model reduces the architectural complexity by collapsing three layers into two. * **Boundary Shift:** In the Hierarchical model, there is a clear visual boundary (the light-grey container) separating the "Metareasoning" layer from the rest of the system. This boundary is absent in the Unified model. * **Information Density:** The Unified model shifts the "Partial Reasoning State" from the reasoning layer down into the representation layer, suggesting a tighter coupling between the state of the reasoning process and the representation of the environment. ### Interpretation The data suggests a fundamental trade-off between **modularity** and **integration**: * **Hierarchical (Modular):** This architecture likely prioritizes separation of concerns. By isolating "Metareasoning," the system can theoretically perform high-level oversight or self-correction without being bogged down by the immediate "Reasoning" or "Situation" processing. It implies a "Controller-Worker" relationship where the controller (Metareasoning) is distinct from the operational layers. * **Unified (Integrated):** This architecture likely prioritizes efficiency and reduced latency. By merging "Reasoning" and "Metareasoning," the system may avoid the overhead of inter-layer communication. Furthermore, by including "Partial Reasoning State" in the "Situation Representation," the system likely aims to create a more holistic "world model" that includes the system's own internal state, potentially allowing for faster, more reactive decision-making. In essence, the Hierarchical model represents a system designed for structured, deliberate control, while the Unified model represents a system designed for fluid, integrated, and potentially more reactive performance. </details> Figure 1: Alternative Metacognitive Architectures. In the CMC and cognitive architectures more generally, a capability is realized through architectural structures and knowledge. Therefore, we propose a unified approach, on the right side of Figure 1, where cognition and metacognition differ only in what is the subject of reasoning. We propose minimal architectural extensions to make information about an agent’s cognition available in working memory. Versions of these extensions are found in existing CMC architectures, including ACT-R [1], Sigma [14], and Soar [8]. We also describe the sources of other non-architectural sources of information about cognition needed to support metacognition as a component of overall cognition. As with the original goals for the CMC, our goal is for our proposal of metacognition to apply to both humans and A(G)I systems with similar general capabilities. Our proposal does not permit information in long-term memories to be examined or reasoned over by other modules. Instead, it restricts reasoning and metareasoning to information available in working memory. It avoids new modules by incorporating the long-term knowledge used in metareasoning within its existing long-term memories. Restricting access to long-term memories sacrifices omniscient metareasoning but enables efficient processing and memory functionalities consistent with neural memory models. Furthermore, all existing cognitive reasoning and learning capabilities are available for metacognition, including interaction with the external environment. Below, we review the Common Model of Cognition and present our proposal for extending it to include human-like metacognition. Our proposal focuses on adding new representational distinctions and sources of information about cognition that become available to an agent to initiate, reason, and terminate metacognition. We attempt to identify the minimum architectural information required to support general metacognition, as well as non-architectural sources of information about cognition that are available to an agent. We include three examples of metacognition within this framework. ## 2 Common Model of Cognition <details> <summary>extracted/6529175/Figures/CMC1.png Details</summary>  ### Visual Description ## Diagram: Cognitive Architecture Model ### Overview The image is a block diagram illustrating a cognitive architecture model. It depicts the flow of information and interaction between various memory systems (Declarative, Procedural, Working), cognitive processing units (Perception, Motor), and the external environment. The diagram serves as a conceptual framework for how an intelligent agent (human or artificial) processes information, retrieves knowledge, and interacts with its surroundings. ### Components/Axes The diagram is composed of six primary functional blocks and two interface modules, connected by directional arrows. **Primary Blocks:** * **Declarative Long-term Memory (Top):** A red rectangular block. * **Working Memory (Center):** A brown rectangular block. * **Procedural Long-term Memory (Left):** A blue rectangular block. * **Perception (Bottom-Left):** A yellow trapezoidal block. * **Motor (Bottom-Right):** A green trapezoidal block. * **Environment (Bottom):** A large purple oval. **Interface Modules:** * **DL (Red box):** Located between Declarative Long-term Memory and Working Memory. * **RL/PC (Blue box):** Located between Procedural Long-term Memory and Working Memory. ### Detailed Analysis The diagram illustrates a closed-loop system with the following flow and connectivity: 1. **Central Processing (Working Memory):** * Positioned centrally. * Acts as the hub for all other components. * **Connections:** * **Top:** Connected to "Declarative Long-term Memory" via a red bidirectional arrow passing through the "DL" interface box. * **Left:** Connected to "Procedural Long-term Memory" via a blue bidirectional arrow passing through the "RL/PC" interface box. * **Bottom-Left:** Connected to "Perception" via a yellow bidirectional arrow. * **Bottom-Right:** Connected to "Motor" via a green arrow pointing downward (from Working Memory to Motor). 2. **Input/Output (Perception & Motor):** * **Perception (Yellow):** Positioned at the bottom-left. It receives input from the "Environment" via a purple arrow pointing upward. It also has a yellow bidirectional arrow connecting it to the "Motor" block. * **Motor (Green):** Positioned at the bottom-right. It sends output to the "Environment" via a green arrow pointing downward. 3. **Environment (Purple):** * Positioned at the very bottom. It serves as the source of input (via the purple arrow to Perception) and the destination for output (via the green arrow from Motor). ### Key Observations * **Centrality of Working Memory:** All cognitive processing and memory retrieval appear to be mediated through the "Working Memory" block, suggesting it is the bottleneck or primary controller of the system. * **Memory Dichotomy:** The diagram explicitly separates long-term memory into two distinct types: "Declarative" (facts/knowledge) and "Procedural" (skills/actions), which is a standard distinction in cognitive psychology. * **Closed-Loop Flow:** The system demonstrates a clear cycle: Environment $\rightarrow$ Perception $\rightarrow$ Working Memory $\rightarrow$ Motor $\rightarrow$ Environment. * **Interface Specialization:** The use of specific interface boxes ("DL" and "RL/PC") suggests that the interaction between Working Memory and Long-term Memory is not direct but requires specific retrieval or processing mechanisms. ### Interpretation This diagram represents a classic cognitive architecture, likely inspired by models such as ACT-R (Adaptive Control of Thought-Rational). * **What the data demonstrates:** It illustrates the functional separation of cognitive tasks. "Perception" and "Motor" handle the interface with the physical world, while "Working Memory" handles the immediate processing. The long-term memory stores are "off-line" repositories that are accessed only when needed by the Working Memory. * **Peircean Investigative Perspective:** * **The "DL" and "RL/PC" labels:** These are likely abbreviations. "DL" likely stands for "Declarative Learning" or "Declarative Link." "RL/PC" likely stands for "Reinforcement Learning / Procedural Control," suggesting that the procedural memory is updated or accessed via these mechanisms. * **The Arrows:** The bidirectional arrows (Red, Blue, Yellow) indicate that information can flow both ways (retrieval and storage/updating). The unidirectional arrows (Green, Purple) indicate a specific flow of control or data (e.g., Perception only feeds into the system; Motor only acts upon the environment). * **System Logic:** The model suggests that an agent cannot act on the environment without first perceiving it and processing that perception through the Working Memory, which is informed by both factual knowledge (Declarative) and skill-based knowledge (Procedural). </details> Figure 2: The Common Model of Cognition. The CMC unifies many similar cognitive architectures by identifying common components, processing, connectivity, and constraints. It does not specify mechanisms or implementations, but emphasizes functionality, focusing on routine cognition and learning. Cognition is the collective processing of the component modules (working memory, procedural memory, declarative memory) and their associated processes (action selection, retrieval, learning). Figure 2 shows the structure and data flow among the modules, which include short-term working memory, long-term procedural memory, long-term declarative memory, learning, perception, and motor control. Long-term memories have associated automatic learning mechanisms that incrementally modify and extend their contents. The CMC posits procedural compilation, reinforcement learning, and declarative learning. Memories contain relations over symbols annotated with quantitative metadata. Examples include the recency of creation or access, probability, and derived utility. Metadata influences processing within a module, such as retrieval and learning. Data flow begins with perception and proceeds to working memory, representing an agent’s understanding of its current situation and goals. Procedural memory contains knowledge about selecting and executing actions that modify working memory. On each cognitive cycle, procedural memory, testing the contents of working memory, selects a single action, which makes one or more changes to working memory, resulting in a step in the cognitive cycle. Each of the other modules has a buffer in working memory through which procedural memory can initiate memory retrievals, motor actions, or top-down control of perception. Results from those module processes are added to their respective buffers. Thus, behavior unfolds as a sequence of steps, driven by procedural memory, making changes to working memory. We call such step-by-step behavior reasoning where the contents of working memory are what the reasoning is about and the knowledge in procedural and declarative memory determines the course of the reasoning. The original CMC encompasses routine and skilled performance, where working memory includes only information about the agent’s current goals and task state. ## 3 Phases of Metacognition Processing In our proposal, metacognition employs the same cognitive cycle process, utilizing the same modules, and often the same knowledge. Metacognition is distinguished by the fact that the information about the agent’s own processing is included in its reasoning: information about prior problem solving (“I see where I made a mistake on that problem and need to do something different in the future.”), specific skills (“I’m good at jigsaw puzzles.”), general competences (“I struggle with math.”), and the operation of individual modules (“I have trouble remembering the difference between affect and effect.”). Our proposal does not introduce new modules but relies on existing architectural structures and knowledge, which are extended to make new forms and representations of information available in working memory. The process of metacognition is not purely metacognition, but a combination of new and existing capabilities. In this section, we describe the three phases of metacognition: initiation, reasoning, and termination. We identify the types of knowledge needed in these phases. Then we describe our proposed extensions to meet those requirements, as well as sources of knowledge available for metareasoning. The following section steps through those phases using three examples of metacognition. ### 3.1 Initiation The standard cognitive cycle for reasoning involves procedural memory responding to changes in working memory, typically concerning the performance of the current task - what we call base-level reasoning. However, a working memory element can be added that indicates a deviation of base-level reasoning, such as a failure in the retrieval from memory. The general requirement is that a knowledge source creates a structure in working memory that is about the agent’s cognition. Such an item can be created deliberately through an action of procedural memory, but also as a side-effect of other long-term memory retrievals or perception. Our proposal extends these by including feedback about the state of individual modules. ### 3.2 Reasoning Once information is available about the state of the agent’s cognition, the agent can use its cognitive capabilities to respond to it (or ignore it), essentially treating it as the signal that a new (meta)problem may need to be solved. Knowledge for reasoning can combine existing task performance knowledge with knowledge that is specific to metacognition. We have identified three extensions to the existing sources of information that can be crucial for metacognition. 1. Information about the current state of agent processing in its modules. 1. A memory of the contents of working memory over time, where a sequence of past situations can be retrieved into working memory and reasoned over. This allows the agent to detect choices it should not make in the future or ones it should reinforce. It also allows the agent to learn a model of the effects of its actions on its internal state, its environment (changes that come from perception), and its processing state. 1. Some means of creating past or future hypothetical states in working memory with two seemingly contradictory properties. One is that these states are represented such that the agent’s knowledge for reasoning in the current state can apply to them, so that the agent can imagine what it would do in those states. The second is that they are also distinguished in some way, so that the agent does not confuse them with reality. Thus, an agent can plan for the future or retrospectively reconsider past actions, without disrupting or interfering with its reasoning about the present. ### 3.3 Termination Metacognition terminates when reasoning is no longer sensitive to representations of the agent’s processing, such as when metacognition resolves the reason it was initiated. Depending on the initiating signal, this can be from a deliberate change to working memory, or indirectly because an impasse in reasoning is resolved. Whatever the reason for termination, any result can also change long-term memory (via the existing learning mechanisms). ## 4 Proposal for Metacognition in the CMC The crux of our proposal is to expand the sources and representations of an agent’s available information to meet the requirements described above. We present our proposal in two stages. First, we describe structural modifications to the CMC that make new representations and direct sources of information available. In the second, we describe how the existing sources can also provide some of that information indirectly. <details> <summary>extracted/6529175/Figures/CMC2.png Details</summary>  ### Visual Description ## Diagram: Cognitive Architecture Model ### Overview This image is a block diagram illustrating a cognitive architecture, likely representing an artificial intelligence agent or a cognitive psychology model. The diagram depicts a central "Working Memory" hub that acts as an interface between various long-term memory systems, sensory perception, motor control, and the external environment. The diagram uses color-coding to link specific memory and processing modules to corresponding interface regions within the central Working Memory block. ### Components/Axes The diagram is composed of several distinct, color-coded functional blocks: * **Central Hub:** * **Working Memory:** A large, brown, horizontally oriented rectangle located in the center of the diagram. It contains smaller, colored "slots" or interface regions that correspond to the external modules. It also features a circular arrow on the right side, indicating an internal feedback loop. * **External Modules (Clockwise from Top-Left):** * **Semantic Long-term Memory:** A maroon rectangle located at the top-left. * **Episodic Long-term Memory:** A red rectangle located at the top-right. * **Motor:** A green rectangle located at the bottom-right. * **Environment:** A large black oval located at the bottom center. * **Perception:** A yellow rectangle located at the bottom-left. * **Procedural Long-term Memory:** A blue rectangle located at the middle-left. * **Interface/Connector Labels:** * **SL:** A maroon box connecting Semantic Long-term Memory to Working Memory. * **EL:** A red box connecting Episodic Long-term Memory to Working Memory. * **RL/PC:** A blue box connecting Procedural Long-term Memory to Working Memory. ### Detailed Analysis #### 1. Connectivity and Flow The diagram utilizes arrows to denote the direction of information flow: * **Semantic Long-term Memory (Maroon):** Connected to Working Memory via a maroon box labeled **SL**. The arrows are bidirectional, indicating a two-way exchange of information. * **Episodic Long-term Memory (Red):** Connected to Working Memory via a red box labeled **EL**. The arrows are bidirectional. * **Procedural Long-term Memory (Blue):** Connected to Working Memory via a blue box labeled **RL/PC**. The arrows are bidirectional. * **Perception (Yellow):** Receives input from the **Environment** (black arrow pointing up). It has a bidirectional relationship with Working Memory (yellow arrows). It also has a bidirectional relationship with the **Motor** module (yellow/green arrow). * **Motor (Green):** Has a bidirectional relationship with Working Memory (green arrows). It sends output to the **Environment** (green arrow pointing down). * **Environment (Black):** Acts as the base of the system. It provides input to Perception and receives output from the Motor module. #### 2. Spatial Grounding of Working Memory Slots The central "Working Memory" block contains internal regions that match the colors of the external modules, suggesting specific "ports" or interface areas: * **Top-Left (Maroon):** Corresponds to Semantic Memory. * **Top-Right (Red):** Corresponds to Episodic Memory. * **Bottom-Left (Yellow):** Corresponds to Perception. * **Bottom-Right (Green):** Corresponds to Motor. * **Far-Left (Blue):** Corresponds to Procedural Memory. ### Key Observations * **Color-Coding Consistency:** The diagram is highly consistent. Every external module has a unique color (Maroon, Red, Blue, Yellow, Green) that is mirrored in the interface connectors and the internal slots of the Working Memory block. * **The "RL/PC" Label:** The Procedural Long-term Memory connector is labeled "RL/PC." In the context of cognitive architectures, this likely stands for **Reinforcement Learning / Procedural Control**, indicating that this module is responsible for skill acquisition and behavioral policy. * **Centrality of Working Memory:** All high-level cognitive functions (Semantic, Episodic, Procedural) and low-level functions (Perception, Motor) must pass through or interact with the Working Memory block. * **Environment Isolation:** The Environment is the only component that does not have a direct connection to the Working Memory block; it is mediated entirely through the Perception and Motor modules. ### Interpretation This diagram represents a classic **Cognitive Architecture** (similar to models like SOAR, ACT-R, or CLARION). * **The "Blackboard" Model:** The Working Memory acts as a "blackboard" or central workspace where different cognitive processes can share information. The internal loop on the Working Memory block suggests a "focus of attention" or "rehearsal" mechanism, where information is maintained or processed iteratively. * **Information Processing Loop:** The flow from Environment -> Perception -> Working Memory -> Motor -> Environment represents the standard "Sense-Think-Act" cycle of an intelligent agent. * **Memory Hierarchy:** The distinction between Semantic (facts), Episodic (events), and Procedural (skills) memories suggests a system designed to handle different types of knowledge. The bidirectional arrows imply that the agent can both retrieve information from these memories and store new experiences or learned skills back into them. * **System Purpose:** The inclusion of "RL/PC" strongly suggests this is an agent designed for autonomous learning, where the agent uses reinforcement learning to update its procedural knowledge based on the outcomes of its actions in the environment. </details> Figure 3: Structural extensions to the CMC to support Metacognition. ### 4.1 CMC Extensions Figure 3 illustrates our proposed extensions. One is that information about the current processing state of each module, including procedural memory and working memory, is available in working memory buffers. The second is the inclusion of the functionality of episodic memory, which is shown as a separation from semantic memory. The third (not shown) is that working memory supports representations of information about past or future states, distinguished from the current state. We also describe how each extension is, or is not, implemented in ACT-R, Sigma, and Soar. #### 4.1.1 Module Process-state Buffers In our proposal, each module has a process-state buffer added to working memory. It summarizes information about the module’s state that can be a signal to initiate metareasoning, as well as the information for understanding the current state of processing during metareasoning. Types of process-state information include the success or failure of requested actions for a given buffer. Additional information can include certainty/confidence in a result, partial results (the answer starts with an ’A’), or indications that an answer is available but not retrieved (feeling of knowing). Perceptual process-state information can include surprise or the inability to recognize parts of the perceptual scene. Process-state information for working memory can include assessments, such as desirability or intrinsic pleasantness. A proposed CMC extension for emotion [15] considers adding a metacognitive assessment module with an associated buffer in working memory, which would be compatible with this proposal. The existing module buffers in ACT-R, Sigma, and Soar include dedicated areas for process-state information. Sigma and Soar include process-state information associated with procedural memory, whereas ACT-R does not. In Sigma and Soar, if procedural memory cannot select an action, a structure called the substate is added to working memory. The substate describes the reason for the impasse and initiates metareasoning and provides a context for metareasoning (see below). Sigma, but not ACT-R or Soar, has something akin to the working memory process-state buffer for representing desirability appraisals. #### 4.1.2 Hypothetical State Representations Our proposal requires that an agent can currently represent hypothetical past or future states with the current state in working memory. The substate mechanisms in Sigma and Soar provide concurrent representations of the current situation for base-level reasoning and hypothetical states and are the locus for metareasoning. The structure of the substates can recreate the current state so that the agent’s long-term knowledge can be used for both metareasoning and base-level reasoning. However, the substates are distinguished by certain features so that hallucination is avoided. ACT-R has no similar architectural mechanisms. Instead, in agents that employ planning, knowledge-based conventions reserve specific slots in their working memory structures (chunks) to distinguish hypothetical states. #### 4.1.3 Episodic Memory Process-state buffers provide information about the instantaneous state of the agent, but do not provide any representation of the history of reasoning that can be used for metareasoning. The ability to reconstruct prior reasoning trajectories is precisely what episodic memory can provide. The original CMC had a single long-term declarative memory. Here, we propose that the functionality associated with episodic memory differs from that of semantic memory, as it provides a direct source of knowledge about an agent’s past reasoning. Episodic learning incrementally and automatically acquires episodes and their temporal relations, allowing an agent to reconstruct the extended sequence of reasoning in working memory, possibly through multiple retrievals. Once in working memory, that data, which is about the agent’s reasoning, enables retrospective analysis and other forms of metareasoning. Soar has distinct semantic and episodic memories as shown in Figure 3; however, ACT-R and Sigma do not. Instead, they have a single long-term declarative memory, where some of the functionalities of episodic memory are supported but not all. ### 4.2 Indirect Sources of Processing Information In addition to the direct sources provided by process-state buffers and episodic memory, an agent can access information about its processing from its interactions with its environment, by metareasoning, from its memories of its reasoning, behavior, and metareasoning. #### 4.2.1 Perception of its Environment: An agent’s environment includes many sources of information that an agent can use to reason about itself. Below are a few general categories. - Self Observation: By observing the results of its interactions with its world, an agent can learn about the impact of its cognitive processes on its environment. When unexpected results occur, the perception of those results can lead to metacognition. When self-observation is combined with episodic memory, an agent can later review its behavior, determining what approaches work and which don’t work, and build up a model for future tasks and associated metareasoning. - Other Agents: Other agents can provide observations about an agent’s cognitive processing, suggest possible reasoning steps or strategies, notify an agent of mistakes (or successes) in its reasoning, or even provide an agent with cognitive strategies (such as through Cognitive Behavior Therapy) to detect its reasoning approaches, evaluate them, and possibly modify them. - Recorded Information: An agent can read books, watch movies, and even study psychology to acquire a general understanding of reasoning capabilities that it applies to itself. Many of these sources provide temporal distance between an agent’s original reasoning and using the information, allowing the agent to create internal representations of behavior and capabilities in working memory after the behavior is generated. This can be impossible during routine reasoning, when only task information is available and task urgency prevents metacognitive processing. #### 4.2.2 Metareasoning: Metareasoning itself composes these other sources of knowledge about an agent’s cognitive capabilities to draw conclusions and generate new insights that fuel future metareasoning. This requires the other originating sources of information, but allows an agent to extend and expand its knowledge by combining information from multiple sources, identifying trends and commonalities, and so on. The knowledge created by metareasoning can then be learned and transferred to long-term memory for future use as described below. #### 4.2.3 Semantic and Procedural Memories: The automatic learning of procedural and semantic knowledge enables the retention of knowledge about an agent’s cognition produced by other sources. They are indirect sources, as they require information to be present in working memory, which, through learning, becomes available for retrieval into working memory and future metacognition. ## 5 Examples of Metacognitive Processing We return to the three phases of metacognitive processing with three examples. ### 5.1 Wordle Retrieval The agent attempts to retrieve a word from semantic memory using a partial specification of a five-letter word. - Initiation: The semantic memory process-state buffer includes information that the retrieved word (‘Tripe’) is uncommon. - Reasoning: The agent decides it wants to know more “about what it knows about the word,” and attempts to retrieve the word from episodic memory, embedding it in the context of it being a Wordle answer. Episodic memory does not retrieve a specific episode of it being a previous answer (possibly because of interference from the hundreds of previous times the agent has played Wordle). Still, the process-state buffer indicates that the word is very familiar. The agent reasons that the combination of being uncommon and familiar indicates it is probably a previous Wordle answer, as it wouldn’t come up in any other situation. - Termination: The agent discards ‘Tripe’ from consideration, which terminates the short bout of metareasoning. The agent returns to generating a potential answer. ### 5.2 Making a Move in Chess The agent is playing a chess game and is far enough into the game that its memorized opening moves are exhausted. Here we describe the mechanisms in Sigma and Soar that support metareasoning in such a situation. - Initiation: Procedural memory does not return a single definite action to take. In Sigma and Soar, this leads to the creation of a substate. - Reasoning: In the substate, the agent decides to explicitly try out the different available moves on internal copies of the current state, generating states that it then compares. - Termination: It ultimately decides on a specific move, which terminates the metareasoning. ### 5.3 Repeated Robot Action A one-armed robot is instructed to store all the leftovers on the counter in the refrigerator. Each time the robot stores an item in the refrigerator, it opens the refrigerator door, fetches the item, places it in the refrigerator, and then closes the refrigerator door. - Initiation: After completing the task, the robot’s instructor tells it that it needs to improve its performance. This external information triggers procedural knowledge that the robot should do a retrospective analysis of its original performance. - Reasoning: Using existing procedural knowledge, the robot recalls a trace of its behavior from episodic memory into working memory. It then uses existing procedural knowledge to analyze the trace. It detects that it is repeatedly in exactly the same world state because it closes the refrigerator door as part of one store command, but then immediately opens it for the next. It “imagines” being at the end of a store command when there are other items to store, and inhibits the action to close the refrigerator door. Through its procedural learning mechanism, it learns to inhibit that action in similar situations in the future. - Termination: On completing its internal inhibition, it has additional procedural knowledge that removes from working memory the comment from the instructor, and continues with its normal activities. ## 6 Conclusion We propose that three specific extensions be added to the CMC to support a unified approach to metacognition: module process-state buffers that provide information in working memory on the current state of each of the buffers; episodic memory that provides a means for the agent to recreate its behavior so that it becomes available to reason over; and some means of creating hypothetical situations in working memory that support using base-level reasoning during metareasoning, avoid interference and confusion between them. Our proposal is a framework, but it does not specify in detail the diverse and extensive indirect learned or pre-encoded long-term knowledge needed for an agent to engage in the forms of metacognition found in humans. An essential point of our proposals is to provide the architectural structure that is necessary above and beyond what is encoded as knowledge in the long-term memories. Although we introduce some new architectural structures, there are no structural boundaries between base-level (non-meta) reasoning and metareasoning. Within existing architectures consistent with this proposal (ACT-R, Sigma, and Soar), an agent can rapidly switch between base-level reasoning and metareasoning, as the only difference is whether working memory structures are about the agent’s cognitive and reasoning capabilities. In addition, an agent’s reasoning transitions from base-level task reasoning to metareasoning as it responds to internal impasses, failures, and successes in base-level reasoning, and then back to base-level reasoning as learning leads to routine, impasse-free reasoning. ## References - [1] Anderson, J.R., Bothell, D., Byrne, M.D., Douglass, S.A., Lebiere, C., Qin, Y.: An integrated theory of the mind. Psychological Review 111 (4), 1036–1060 (2004) - [2] Cox, M., Alavi, Z., Dannenhauer, D., Eyorokon, V., Munoz-Avila, H., Perlis, D.: MIDCA: A Metacognitive, Integrated Dual-Cycle Architecture for Self-Regulated Autonomy. Proceedings of the AAAI Conference on Artificial Intelligence 30 (1) (Mar 2016) - [3] Cox, M., Raja, A.: Metareasoning: An Introduction, pp. 3–14. MIT Press (2011) - [4] Flavell, J.H.: Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. American Psychologist 34 (10), 906–911 (1979) - [5] Johnson, S.G.B., Karimi, A.H., Bengio, Y., Chater, N., Gerstenberg, T., Larson, K., Levine, S., Mitchell, M., Rahwan, I., Schölkopf, B., Grossmann, I.: Imagining and building wise machines: The centrality of AI metacognition (2025), https://arxiv.org/abs/2411.02478 - [6] Kotseruba, I., Tsotsos, J.K.: 40 years of cognitive architectures: Core cognitive abilities and practical applications. Artificial Intelligence Review 53 (1), 17–94 (Jan 2020) - [7] Kralik, J.D., Lee, J.H., Rosenbloom, P.S., Jackson, P.C., Epstein, S.L., Romero, O.J., Sanz, R., Larue, O., Schmidtke, H.R., Lee, S.W., McGreggor, K.: Metacognition for a Common Model of Cognition. Procedia Computer Science 145, 730–739 (Jan 2018) - [8] Laird, J.E.: The Soar Cognitive Architecture. MIT Press, Cambridge, MA (2012) - [9] Laird, J.E., Lebiere, C., Rosenbloom, P.S.: A Standard Model of the Mind: Toward a Common Computational Framework across Artificial Intelligence, Cognitive Science, Neuroscience, and Robotics. AI Magazine 38 (4), 13–26 (2017) - [10] Laird, J.E., Mohan, S.: Learning Fast and Slow. In: Proc. of the 32nd AAAI Conference on Artificial Intelligence. p. 5. AAAI Press, New Orleans (2018) - [11] Nelson, T.O., Narens, L.: Metamemory: A theoretical framework and new findings. In: Bower, G. (ed.) The Psychology of Learning and Motivation: Advances in Research and Theory, pp. 125–173. Academic Press, New York (1990) - [12] Nolte, R., Pomarlan, M., Janssen, A., Beßler, D., Javanmardi, K., Jongebloed, S., Porzel, R., Bateman, J., Beetz, M., Malaka, R.: How metacognitive architectures remember their own thoughts: A systematic review (2025), https://arxiv.org/abs/2503.13467 - [13] Rahnev, D., Balsdon, T., Charles, L., de Gardelle, V., Denison, R., Desender, K., Faivre, N., Filevich, E., Fleming, S.M., Jehee, J., Lau, H., Lee, A.L.F., Locke, S.M., Mamassian, P., Odegaard, B., Peters, M., Reyes, G., Rouault, M., Sackur, J., Samaha, J., Sergent, C., Sherman, M.T., Siedlecka, M., Soto, D., Vlassova, A., Zylberberg, A.: Consensus goals for the field of visual metacognition. Perspectives on Psychological Science 17 (6), 1746–1765 (2022) - [14] Rosenbloom, P.S., Demski, A., Ustun, V.: The Sigma Cognitive Architecture and System: Towards Functionally Elegant Grand Unification. Journal of Artificial General Intelligence 7 (1), 1–103 (Dec 2016) - [15] Rosenbloom, P.S., Laird, J.E., Lebiere, C., Stocco, A., Granger, R., Huyck, C.: A proposal for extending the Common Model of Cognition to emotion. In: 22nd International Conference on Cognitive Modeling, ICCM 2024. Tilburg University, the Netherlands (Apr 2024) - [16] Sun, R.: The importance of cognitive architectures: an analysis based on CLARION. Journal of Experimental & Theoretical Artificial Intelligence 19 (2), 159–193 (2007) - [17] Walker, P., Haase, J., Mehalick, M., Steele, C., Russell, D., Davidson, I.: Harnessing metacognition for safe and responsible ai. Technologies 13 (2025)