A Computational Model of
Reflexive and Reflective Reasoning
Integrating Reflexive and Reflective Processes

The immediate goal of this work was to develop a tool that:

  1. Performs rapid recognitional inferences and planning within a large (expert) belief network
  2. Exemplifies human limitations on computational resources and attention
  3. Uses reflective strategies to help overcome computational limitations and deal with uncertainty.

Such a tool would be capable of simulating highly proficient, subtle, and creative aspects of human decision making in real-world domains.



The Reflexive System

Proficient decision makers, in relatively familiar situations, rapidly settle on situation interpretations and plans in the face of new observations and changing goals. This kind of situaion understanding and planning can be largely understood in terms of causal knowledge structures that we call mental models.

The reflexive subsystem shows how mental models can be generated and maintained in a dynamic situation. The starting point of the reflexive subsystem was a system called Shruti, developed by Professor Lokendra Shastri, of the International Computer Science Institute, University of California at Berkeley (Shastri & Ajjanagadde, 1993). Dr. Shastri is a long-time collaborator of CTI.

Shruti is unique in combining speed and scalability, with representation of crucial relational aspects of real-world decision making. To accomplish speed and scalability, Shruti utilizes parallel, connectionist processing. To keep track of relational reasoning, Shruti uses temporal synchrony of firing in nodes throughout the network to represent information about the same object. (Most connectionist models represent fuzzy similarity relations that blur the way objects interact. Symbolic architectures easily represent relational facts, but are usually slow and non-scalable.)

In Shruti, activation goes out from nodes that represent either sensory or linguistic inputs and/or an internally generated question to be settled, and returns when circuit is completed by other nodes that are linked to them. Shruti represents a situation model and/or decision by the emergence of a stable activation cycle within the network.

Extensions to Shruti were necessary both to improve its representation of reflexive reasoning and to make it work in conjunction with a reflective subsystem. Among the extensions that we have developed are:

  1. Propagation of utility as well as belief, so that Shruti settles on actions at the same time as it settles on a situation interpretation.
  2. Reasoning both backward, to find explanations, and forward, to generate predictions.
  3. Mechanisms required for shifting attention, such as (a) priming effects that temporarily store results through a series of attentional shifts, and (b) long-term storage of aggregated results at the edge of the currently active part of the network.
  4. Tuning of link strengths through backpropagation.

Attentional limits on dynamic access to long term memory (LTM) emerge naturally from the computational structure of Shruti and the neuro-biological constraints it respects. These inmply that not all information known by the agent can be brought to bear on a problem at the same time by purely reflexive processes.


The Reflective System

The reflective subsystem critiques the conclusions of reflexive processing and guides its subsequent progress. (See the Recognition / Metacognition model) Features of the reflective subsystem include:

  1. Methods for identifying qualitatively different types of uncertainty based on activation patterns in the reflexive system
  2. Methods for identifying beliefs most likely to be responsible for different types of uncertainty
  3. Strategies for shifting attention to beliefs most likely to be responsible for uncertainty. The results of these attentional shifts is the activation of previously dormant information in long-term memory that is likely to help resolve the uncertainty.

The metacognitive system learns to combine a set of simple operations: inhibiting recognitional responding, activation of new information internally by shifting focal attention, and clamping truth values (which is itself a form of persistent attention to a node). These operations are simple and both psychologically and biologically plausible.

The metacognitive system learns to combine these operations in response to different patterns of uncertainty by reinforcement and associative learning processes. Through such learning, the metacognitive system acquires a rich repertoire of uncertainty handling strategies. These strategies include both domain-specific and general elements. The result is a dynamic process of evaluating and improving mental models of the situation and plans.


Summary
In sum, there are inherent, and dynamic, limits on the scope of LTM information that can be brought to bear in interpreting evidence or answering questions. The existence of such limits means that inference and planning processes must be capable of (a) dynamically determining the scope of active human memory from which they draw at any given time, and (b) of remaining coherent within those limits. This need for fluid changes in focus introduces the necessity for an adaptive dynamics of executive attention. The key interaction between the reflexive and reflective systems is the adaptive direction of focused attention within the reflexive memory by means of learned metacognitive behaviors. Recency effects are used to assemble such intermediate results into composite assessments. The model suggests that the development of executive attention functions (metacognitive strategies) may be necessary for, and integral to, the development of working memory, or dynamic access to LTM.


Funded projects
A Hybrid Architecture for Metacognitive Learning

The initial funding for this research was conducted under a program entitled "Hybrid Architectures for Complex Learning." The research was funded by the Office of Navy Research as part of the Hybrid Architectures for Complex Learning program and there is a web site for all of the research projects funded by this program.

The focus of the ONR program was on hybrid combinations of AI and connectionist techniques. We were funded for a proposal in which a classically "AI" component, an inferential memory, was implemented as a reflexive reasoner in a connectionist network (and coupled to a distributed connectionist "metacognitive" controller in the proposal). The term "hybrid" has stuck -- perhaps it now refers mainly to the hetero-duality of the recognitional and metacognitive components which we use to model naturalistic decision making.

In this project, we began with a cognitive model of recognitional and metacognitive behaviors and translated this model into a connectionst architecture. This research was done in close collaboration with Lokendra Shastri and uses Shruti to implement the recognitional behaviors.

Mental Models This project is concerned with how skilled decision makers assemble and act on situation estimates developed from mental models.  The primary focus of this work has been training critical thinking skills.  In conjunction with that research, we have been exploring computational models of both the recognitional skills by which people reflexively elaborate explanations and predictions in coherent stories, and the metacognitive skills which experts develop to critique and interatively improve those initial assessments.  These computational models may be applied in several ways, including:
  1. intelligent training system, including recommending material and exercises and evaluating user responses
  2. to facilitate information management and help train sophisticated metacognitive filtering strategies.
This work is funded by The U.S. Army Research Institute for the Behavioral and Social Sciences -- see the Mental Models page for more information on this project.


See also

 

Copyright © 2000-2011 Cognitive Technologies, Inc.
Questions? Comments? Contact webmaster@cog-tech.com