Patrick Simen Research Fellow Princeton Neuroscience Institute Princeton University Green Hall Washington Rd. Princeton, NJ 08544-1000 Phone: (609) 258-5032 psimen@math.princeton.edu

Peer-reviewed journal and conference publications

• Simen, P. and Polk, T. (in press). A symbolic/subsymbolic interface protocol for cognitive modeling, Logic Journal of the Interest Group in Pure and Applied Logic (IGPL). Supplement.
• Mulder, M. J., Gold J. I., Durston, S., Heasly, B., Millner, A., Simen, P., Getz, S., Voss, H., Ballon, D. and Casey, B. J. (in review). BOLD correlates of reward-related decision bias on a visual discrimination task.
• Simen, P., Contreras, D., Buck, C., Hu, P., Holmes, P. and Cohen, J. D. (in press). Reward-rate optimization in two-alternative decision making: empirical tests of theoretical predictions, Journal of Experimental Psychology: Human Perception and Performance.
• Simen, P. and Cohen, J. D. (2009). Explicit melioration by a neural diffusion model, Brain Research, 1299:95-117.
• Gao, J., Wong-Lin, K.F., Holmes, P., Simen, P. and Cohen, J. D. (2009). Sequential effects in two-choice reaction time tasks: decomposition and synthesis of mechanisms, Neural Computation, 21:2407-2436.
• Simen, P., Cohen J. D., and Holmes, P. (2006). Rapid decision threshold modulation by reward rate in a neural network (.pdf), Neural Networks, 19:1013-1026.
• Simen, P., Polk, T., Lewis, R., and Freedman, E. (2004). A computational account of latency impairments in problem solving by Parkinson's patients (.pdf), Proceedings of the International Conference on Cognitive Modeling, 273-279.
• Simen, P., Polk, T., Lewis, R., and Freedman, E. (2003). Universal computation by networks of model cortical columns (.pdf), Proceedings of the International Joint Conference on Neural Networks, July 2003, 230-235.
• Polk, T., Simen, P., Lewis, R., and Freedman, E. (2002). A computational approach to control in complex cognition (.pdf), Cognitive Brain Research, 15:71-83.
• Simen, P., Polk, T., Lewis, R., and Freedman, E. (2002). A recurrent neural network model of goal management (.pdf), Proceedings of the International Conference on Computational Intelligence and Neuroscience, 504-508.

Manuscripts in Preparation

• McMillen, T., Simen, P. and Behseta, S. (in preparation). Reward-modulated Hebbian learning leads to near-optimal performance in decision making tasks with more than two alternatives.
• Simen, P. and Balci, F. (in preparation). Adaptive interval timing by a noisy integrate-and-fire model.
• Balci, F., Simen, P., Niyogi, R., Holmes, P. and Cohen, J. D. (in preparation). Acquisition of optimal speed-accuracy tradeoffs by humans.

Technical Reports

• Simen, P., Freedman, E., Lewis, R., and Polk, T. (2003). Columnar timing mechanisms in neural models of problem solving (.pdf), Computer Science and Engineering technical report CSE-TR-481-03.

Book chapters

• Simen, P. (in press). Decision making and reward: computational perspectives. Encyclopedia of Mind, Pashler, H. (ed.), SAGE Publications.
• Simen, P., Holmes, P. and Cohen, J. D. (2008). On the neural implementation of optimal decisions, Oxford Handbook of Human Action, Morsella, E., Bargh, J. A. and Gollwitzer, P. M. (eds.), Oxford University Press, 534-549.

Talks/Posters

• Simen, P., Nystrom, L., Van Vugt, M., Sederberg, P., Balci, F. and Cohen, J. D. (2009). Event-related fMRI during slow decision making can reveal temporal structure in neural activity, poster presented at the 2009 meeting of the Society for Neuroscience.
• Simen, P., Contreras, D., Holmes, P. and Cohen, J. D. (2009). Adaptive performance in two-alternative decision making. Talk presented at the 2009 meeting of the Society for Mathematical Psychology.
• Simen, P., Contreras, D., Buck, C., Hu, P., Holmes, P. and Cohen, J. D. Reward-maximizing performance in two-alternative decision making, poster presented at the 2008 meeting of the Psychonomic Society.
• Simen, P. Ramping, ramping everywhere: an overlooked model of interval timing, poster presented at COSYNE 2008.
• Simen, P. and Cohen J. D., A diffusion-based neural network model of interval timing and temporal discounting, poster presented at the 2007 Society for Neuroscience Conference.
• Simen, P. and Cohen, J. D., Explicit melioration by a simple neural network, poster presented at the 2007 Computational Cognitive Neuroscience/Dynamical Neuroscience Conference.
• Simen, P., Holmes, P. and Cohen, J. D., Melioration by a diffusion model with response threshold adaptation, poster presented at the 2006 Society for Neuroeconomics conference.
• Simen, P., Holmes, P. and Cohen, J. D., Threshold adaptation in decision making, poster presented at the 2005 Society for Neuroscience conference.
• Simen, P., Holmes, P. and Cohen, J. D., A model of threshold adaptation in perceptual decision making, poster presented at the 2005 Cognitive Neuroscience Society meeting.
• Simen, P., Freedman, E., Lewis, R. and Polk, T., Sequence learning in model frontostriatal circuits, poster presented at the 2003 Society for Neuroscience meeting.
• Simen, P., Polk, T., Lewis, R. and Freedman, E., Modeling executive control, problem-solving and sequencing in neural networks (.pdf), poster presented at the 2003 annual meeting of the Cognitive Neuroscience Society.
• Simen, P., Polk, T., Lewis, R., and Freedman, E., A recurrent neural network model of executive control in the Tower of London task (.pdf), poster presented at the 2002 annual meeting of the Cognitive Neuroscience Society.

PhD thesis

• Neural mechanisms for control in complex cognition
  2004, Computer Science and Engineering, University of Michigan.

Curriculum Vitae

Research

I am currently a postdoctoral research fellow at the Princeton Neuroscience Institute (PNI), Princeton University, working with Jonathan D. Cohen in Psychology (co-director of PNI) and Philip Holmes in the Program in Applied and Computational Mathematics and the Department of Mechanical and Aerospace Engineering. The focus of our research is on the role of reward and cost monitoring in human and animal performance of decision-making tasks. We attempt to model basic decision-making circuits in the brain, using dynamical systems models that are as simple as possible, but that achieve enough functionality to account for features of both behavioral and physiological data. Our empirical work focuses specifically on neural mechanisms for controlling decision-making circuits and optimizing their performance in terms of reward maximization. We test the predictions of our models of these mechanisms with experiments in human behavior, EEG and fMRI.

Before my postdoctoral work, I received a Ph.D. in Computer Science and Engineering, focusing on Intelligent Systems, from the University of Michigan. There I worked with Thad Polk in Psychology, and in collaboration with Rick Lewis in Psychology and Eric Freedman in Psychology at University of Michigan, Flint. We used recurrent neural networks to create computational models of human problem solving in the Tower of London task (a psychological task much like the disk-stacking Tower of Hanoi task). The Tower of London task is a good task for identifying problem-solving impairments in patients who have suffered prefrontal cortical damage and patients with Parkinson's disease, which is primarily a disease of the basal ganglia. The cognitive functions of these brain areas and the functional interactions between them were of particular interest to us, because these areas appear to subserve the sort of symbolic processing that distinguishes classic cognitive psychological and artifical intelligence models from the less structured network models that are intended to model the brain in more physical detail.