The Soar Cognitive Architecture

By John Edwin Laird. 2012

1,119 citations recorded by the source corpus.

Subfields: Mechanical Engineering, Knowledge Graphs & Representation, Cognitive Science & Psychology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Mechanical Engineering

In Knowledge Graphs & Representation

In Cognitive Science & Psychology

Citations in the corpus

Explainable Multi-hop Verbal Reasoning Through Internal Monologue (2021, methodology): “, 1997) and SOAR (Laird, 2012), which originate from Newell’s GPS.”
Building Cooperative Embodied Agents Modularly with Large Language Models (2023, methodology): “Inspired by prior work in cognitive architectures (Laird, 2019), we present CoELA , a Co operative E mbodied L anguage A gent, a cognitive architecture with a novel modular framework that utilizes the rich world knowledge, strong reasoning ability and mastery natural language understanding and…”
Executive functions and the generation of “random” sequential responses: A computational account (2016, methodology): “Critically, the Supervisory System part of the model draws also on the executive functions investigated by Miyake et al. (2000) and the decomposition of supervisory processes outlined by Shallice and Burgess (1996). Subsequent sections report simulation studies that demonstrate that the model can ca”
Artificial General Intelligence: Concept, State of the Art, and Future Prospects (2009, background): “• SOAR (Laird, 2012), a classic example of expert rule-based cognitive architecture. designed to model general intelligence.”
DREAM Architecture: a Developmental Approach to Open-Ended Learning in Robotics (2020, positive): “Most traditional cognitive architectures (e.g. ACT-R, CLARION, SOAR) are based on symbolic representations and, thus, implement relatively straightforward LTM structures where reuse is based on searching for the appropriate labels.”
40 years of cognitive architectures: core cognitive abilities and practical applications (2018, methodology): “Image data is sub-symbolic inMAMID (Reisenzein et al. 2013) and SASE (Weng 2002) and non-symbolic (iconic) in Soar (Laird 2012b) and RCS (Albus and Barbera 2005).”
Generative Agents: Interactive Simulacra of Human Behavior (2023, background): “The architectures used by these agents difered (Quakebot-and TacAir-SOAR relied on SOAR [61], while ICARUS relied on its own variation that was inspired by SOAR and ACT-R [6]), but they shared the same underlying principle [62].”
A Standard Model of the Mind: Toward a Common Computational Framework across Artificial Intelligence, Cognitive Science, Neuroscience, and Robotics (2017, background): “However, the model is agnostic as to whether symbols are uninterpreted labels, such as in Lisp, Soar (Laird 2012), and ACT-R (Anderson 2007), or whether they are patterns over vectors of distributed elements, such as semantic pointers in Spaun (Eliasmith et al.”
A Survey on the Memory Mechanism of Large Language Model-based Agents (2024, neutral): “We summarize previous works on memory operations in Table 3.”
On Defining Artificial Intelligence (2019, methodology): “The initial design of Soar attempted to meet the needs of both artificial intelligence and cognitive psychology (Newell, 1990), while in the recent years the project has been mainly shaped by AI considerations (Laird, 2012).”
AI Agents vs. Agentic AI: A Conceptual Taxonomy, Applications and Challenges (2025, positive): “Systems like SOAR [11] and the subsumption architecture [12] extended symbolic and reactive logic into cognitive modeling and robotics.”
Robots That Use Language (2020, background): “3 language commands given to a robotic forklift, as well as to interpret route instructions for a wheelchair (14 language by mapping to planning formalisms (75), an approach that dramatically increases the speed with which words can be interpreted by the robot .”
Nengo: a Python tool for building large-scale functional brain models (2014, background): “…Brette, 2008; Eppler et al., 2009; Hines et al., 2009)], and many tools for simulating cognitive phenomena with various levels of biologically plausibility [e.g., (Cooper and Fox, 1998; Sun, 2001; Anderson et al., 2004; Franklin et al., 2007; Aisa et al., 2008; de Kamps et al., 2008; Laird, 2012)].”
Machine learning \& artificial intelligence in the quantum domain (2017, background): “Another perspective tends to try to capture cognitive architectures, such as SOAR developed by John Laird, Allen Newell, and Paul Rosenbloom (Laird, 2012).”
Optimal Behavioral Hierarchy (2014, background): “In Soar, subgoals specifically related to solving impasses in decision making are automatically acquired.”
Cognitive Architectures for Language Agents (2023, background): “We next provide a brief overview of Soar and refer readers to Laird (2022; 2019) for deeper introductions.”