Home About DANs Projects Past Activity Team Members Contact
Current Projects and Papers in Progress
Book Project: Dynamic Associative Networks and the Foundations of Network Computing. (Outline). Author: Anthony Beavers.
Modeling Project: Inter-generational Learning and the Emergence of Culture in AI Agents. Leads: Paul Eugenio and Anthony Beavers. Student Modeler: Ryan Cerauli.
Modeling Project: Using Dynamic Associative Networks with Financial Data as a Test for Market Efficiency. Leads: Anthony Beavers and Omer Bayar. Student Modeler: Oren Stewart. Business Consultants: Jeff Brosman and Hiten Sonpal.
Conceptual Paper: Neuromorphic Computing, Neuroscience, and Phenomenology. Co-Authors: Anthony Beavers, Denizhan Pak, Coutrney Rowe, Ling Sun, Alex Webb, and Matt Winter.
Conceptual Paper: Bridging AI, Phenomenology, and Consciousness: The Challenge of Artificial Subjectivity. Authors: Steven Gouveia and Anthony Beavers. Invited for the Journal of Artificial Intelligence and Consciousness. (Call for Papers).
Conceptual Paper: The Importance of Intersubjective Communication for Information Novelty. Authors: Devin Wright and Anthony Beavers.
Undergraduate Research Projects - Fall 2025
Ryan Cerauli. Inter-generational Learning and the Emergence of Culture in AI Agents. Full team listed above. Ryan is using a gamified world model architected by Paul Eugenio to study under what conditions DANs might self-assemble.
Ryan Cerauli and Courtney Rowe. Exploring the Relationship between DANs and ANNs. Ryan is studying ways in which DANs might be compressed for greater computational efficiency and whether such compression turns DANs into ANNs. Courtney is studying whether benign overfitting with ANNs captures the same clusters as those in equivalent DANs.
Nathan Stergar. Exploring Possibility and Probability in Dynamic Associative Networks. Nathan is exploring the way that possibility and probability are represented in DANs and how each (or a combination of each) might best contribute to network activation.
Oren Stewart. Using Dynamic Associative Networks with Financial Data as a Test for Market Efficiency. Full team listed above. Oren is rebuilding DAN Core Content-Addressable Memory in Python in order to work with very large vectors for financial market prediction.
Recent Projects and Papers
Student Project: Using a DAN to Segment Words in a Random String of Text, Spring 2025. Student Modelers: Amogh Busner, Ryan Cerauli, Nathan Stergar, and Oren Stewart. DAN model based on P. Myles Eugenio, Hebbian Learning the Local Structure of Language, 2025. This model showed decreasing entropy as words were segmenting with a sharp and sudden increase in entropy at word endings as predicted. This indicates that entropy measures can be used to segment information into temporally occurring patterns in a DAN.
Colloquium Presentation: Network Idealizations and the Prospect of an Associationist Phenomenology. Presenter: Anthony Beavers - Department of the History and Philosophy of Science and Medicine, Indiana University, January 16th, 2025 (PDF of PowerPoint Slides).
Colloquium Presentation: On the Foundations of Network Computing for Artificial Intelligence Purposes. Presenter: Anthony Beavers - Logic Seminar, Department of Mathematics, Indiana University, November 20th, 2024 and December 4th, 2024 (PDF of PowerPoint Slides).
Conceptual Paper: Clues and Caveats concerning Artificial Consciousness from a Phenomenological Perspective. Authors: Anthony Beavers and Eli B. McGraw. Phenomenology and the Cognitive Sciences 23.5 (2024): 1073-1095. Introduces the notion of "associationist phenomenology" suggesting that associative mechanisms operating on qualia defined as "somatically-experienced sense data" might account in part for consciousness. (While DANs are the associative mechanisms imagined in the paper, they are not explicitly discussed here.)
Past DAN-Related Projects Done at the University of Evansville
Open Trading Poker, Fall 2020. Students: Austin Davidson, Angela Jansen, Jacek Knaur, Nick Lewis, and Alec Votoupal. Developed an agent-based model to test various trading heuristics to explore how they affect the distribution of wealth in an artificial society. The significance of the project for DANs was to explore artificial decision making networks that might respond to heuristics rather than thresholds embedded in network nodes.
Predicting the NFL, Spring 2015-Spring 2016. Students: Blake Adams, Cody Baker, Jacob Ball, Andrey Biryuchinskiy, Chris Cannon, Adam Devery, Kiefer Goldman, Jacob Green, Riley Mayr, Trevor Mullen, Craig Schlemmer, Evan Snider, and Daniel Waskiewicz. Built DAN-based content-addressable memory out of fifteen years of NFL data and then used that memory to predict winning teams in a novel season. Demonstrated that DAN-based content-addressable memory can perform predictive inference by association.
Transforming Standard Databases into Teleodynamic and Predictive Mechanisms, Summer 2011. Student: Christopher Harrison. Supported by the UExplore Undergraduate Research Program, The University of Evansville. Final paper published as Beavers, A., and Harrison, C. “Information-Theoretic Teleodynamics in Natural and Artificial Systems.” In A Computable Universe: Understanding Computation & Exploring Nature as Computation, edited by Hector Zenil (World Scientific, 2012), 347-364. First steps toward transducing a database into a partially-connected directed graph for guided search purposes. DAN core memory models are basically isomorphs of the distribution of data in a dataset rendered as partially-connected directed graphs. This will not be clear in the aforementioned paper, though it nonetheless sowed the seeds for how current DAN core memory models are architected.
Dynamic Associative Network Automatic Document Classification. Two year project: 1) 2009-2010. Student: Guy Wyant. Winner of the University of Evansville's Outstanding Senior Project in Computer Science Award, 2010. 2) 2010-2011. Student: Derek Burrows. Paper presented at the Undergraduate Conference in Math, Engineering, and Science, University of Evansville, March 26th, 2011. Winner of the University of Evansville's Outstanding Senior Project in Computer Science Award, 2011. Supported by the National Endowment for the Humanities. This project used dynamic associative networks to extract semantic signatures from articles in the Stanford Encyclopedia of Philosophy that could, in turn, be used to classify novel philosophical documents.
Noesis: Philosophical Research Online and the Indiana Philosophy Ontology (InPho) Project, Spring 2010-Summer 2011. Students: Mason Blankenship, Christopher Harrison, Justin Simerly, and Joshua Woody. Supported by the National Endowment for the Humanities. Developed an interface between Colin Allen's emergent taxonomy of the discipline of philosophy and the Noesis search engine. For more on the InPho project, see Buckner, C., Niepert, M., and Allen, C. From Encyclopedia to Ontology: Toward Dynamic Representation of the Discipline of Philosophy. Synthese 182.2 (2011): 205-233. For more on the Noesis search engine, see below.
Tackling McClelland and Rumelhart's Interactive Activation Model with a Dynamic Associative Network, 2010. Modeler: Anthony Beavers. This was a DAN rebuild of the Jets and Sharks model in McClelland, J. L. Retriving General and Specific Information from Stored Knowledge of Specifics. Proceedings of the Third Annual Conference of the Cognitive Science Society. Berkeley, California, August 19-21, 1981. Results presented as More Fun with Jets and Sharks: Typicality Effects and the Search for the Perfect Attractors. North American Meeting of the International Association for Computing and Philosophy, Simulations and Their Philosophical Implications, Carnegie Mellon University, July 24th-26th, 2010. Results published as Typicality Effects and Resilience in Evolving Dynamic Networks. In FS-10-03, The Association for the Advancement of Artificial Intelligence Press, 2010.
Modeling and Visualizing Dynamic Associative Networks: Towards Developing a More Robust and Biologically Plausible Cognitive Model, 2008-2009. Student: Michael Zlatkovsky. Paper presented at the Undergraduate Conference in Math, Engineering, and Science, University of Evansville, April 4th, 2009. Winner of the University of Evansville's Outstanding Senior Project in Computer Science Award, 2009. This project implemented DANs in a Java application with a graphical visualizer that showed how the graph changed as new information was consolidated into the network.
Modeling Primary Sequential Memory with a Dynamic Associative Network, 2008. Modeler: Anthony Beavers. This was a DAN rebuild of the model reported in Allen, C., and Lange, T. Primary Sequential Memory: An Activation-Based Connectionist Model. Neurocomputing 11:2-4 (1996): 227-243. This was our first attempt to control the flow of information in a DAN to support temporal representations. In this case, both models receive a seven digit number and then return it digit by digit over time in the order in which it was received.
Quality-Controlled Search Engine Work
Noesis: Philosophical Research Online. Launched in 1998. (PDF of the Noesis About Page from 2000). Co-editors: Anthony Beavers and Peter Suber. Students: Chris Batson, Mason Blankenship, Josh Burger, Derek Burrows, Jeff Carlyle, Scott Glenn, Christopher Harrison, Trenton Kriete, Brian Moffatt, Siddartha Naidu, Jason Schindler, Justin Simerly, Hiten Sonpal, Joshua Woody, and Guy Wyant. There were several publications and presentations on the Noesis project. The last and most authoritative statement is Beavers, A. Noesis and the Encyclopedic Internet Vision. Synthese 182.2 (2011): 315-333.
Hippias: Limited Area Search of Philosophy on the Internet. Launched in 1997. (PDF of the Hippias About Page from 1997). Editor: Peter Suber. Managing Editor: Anthony Beavers. Software Architect: Hiten Sonpal. Hippias was a recapulation of Argos using the same protocol to limit searchable resources to the discipline of philosophy.
Argos: Limited Area Search of the Ancient and Medieval World. Launched in 1996. (PDF of the Argos About Page from 1997). Editor: Anthony Beavers. Software Architect: Hiten Sonpal. Argos was the first search engine on the Internet with emergent peer review. Quality was controlled by constraining the searchable area of the Internet by shutting down pathways on a directed graph based on human editorial decisions. See Jeffrey Young, Professor Develops 'Peer-Reviewed Search Engine'. Chronicle of Higher Education: Academe Today, October 10th, 1996. Also in Campus Review (Australia), October, 1996. Argos was also the subject of articles in the Washington Post and the New York Times.
Copyright © 2025. The DAN Research Group. All Rights Reserved.