SAL: A Hybrid Cognitive Architecture Y. Vinokurov 1, C. Lebiere 1, D. Wyatte 2, S. Herd 2, R. O’Reilly 2 1. Carnegie Mellon University, 2. University of.

Slides:



Advertisements
Similar presentations
A Comparison of Rule-Based versus Exemplar-Based Categorization Using the ACT-R Architecture Matthew F. RUTLEDGE-TAYLOR, Christian LEBIERE, Robert THOMSON,
Advertisements

Ch. Eick: More on Machine Learning & Neural Networks Different Forms of Learning: –Learning agent receives feedback with respect to its actions (e.g. using.
Introduction to Training and Learning in Neural Networks n CS/PY 399 Lab Presentation # 4 n February 1, 2001 n Mount Union College.
A model of Consciousness With neural networks By: Hadiseh Nowparast.
Artificial Neural Networks - Introduction -
Artificial Neural Networks - Introduction -
Faculty of Management and Organization Emergence of social constructs and organizational behaviour How cognitive modelling enriches social simulation Martin.
Chapter 10 Artificial Intelligence © 2007 Pearson Addison-Wesley. All rights reserved.
Content Based Image Clustering and Image Retrieval Using Multiple Instance Learning Using Multiple Instance Learning Xin Chen Advisor: Chengcui Zhang Department.
Simple Neural Nets For Pattern Classification
Laurent Itti: CS599 – Computational Architectures in Biological Vision, USC Lecture 6: Low-level features 1 Computational Architectures in Biological.
Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural.
ACT-R.
Psychology of Music Learning Miksza Memory Processes.
Instar Learning Law Adapted from lecture notes of the course CN510: Cognitive and Neural Modeling offered in the Department of Cognitive and Neural Systems.
DDMLab – September 27, ACT-R models of training Cleotilde Gonzalez and Brad Best Dynamic Decision Making Laboratory (DDMLab)
November 30, 2010Neural Networks Lecture 20: Interpolative Associative Memory 1 Associative Networks Associative networks are able to store a set of patterns.
CIS465: Adv. Info. Sys. Memory Systems Min Song Information Systems Dept.
Laurent Itti: CS599 – Computational Architectures in Biological Vision, USC Lecture 7: Coding and Representation 1 Computational Architectures in.
Face Recognition Using Neural Networks Presented By: Hadis Mohseni Leila Taghavi Atefeh Mirsafian.
Economic Attention Networks: Associative Memory and Resource Allocation for General Intelligence Matthew Iklé, Joel Pitt, Ben Goertzel, George Sellman.
Machine Learning. Learning agent Any other agent.
Proceedings of the Twenty-Sixth AAAI Conference on Artificial Intelligence Using Expectations to Drive Cognitive Behavior Unmesh Kurup, Christian Lebiere,
A Recommendation Mechanism for Contextualized mobile advertising S.-T. Yuan et al., Expert Systems with Applications, vol. 24, no. 4, pp , 2003.
Chapter 11: Artificial Intelligence
Modeling Driver Behavior in a Cognitive Architecture
Presentation on Neural Networks.. Basics Of Neural Networks Neural networks refers to a connectionist model that simulates the biophysical information.
Module 12 Remembering & Forgetting. INTRODUCTION recall –retrieving previously learned information without the aid of or with very few external cues recognition.
Slides are based on Negnevitsky, Pearson Education, Lecture 12 Hybrid intelligent systems: Evolutionary neural networks and fuzzy evolutionary systems.
Artificial Neural Nets and AI Connectionism Sub symbolic reasoning.
Self organizing maps 1 iCSC2014, Juan López González, University of Oviedo Self organizing maps A visualization technique with data dimension reduction.
Beyond Gazing, Pointing, and Reaching A Survey of Developmental Robotics Authors: Max Lungarella, Giorgio Metta.
Introduction to ACT-R 5.0 ACT-R Post Graduate Summer School 2001 Coolfont Resort ACT-R Home Page: John R. Anderson Psychology Department.
Hebbian Coincidence Learning
Boltzmann Machine (BM) (§6.4) Hopfield model + hidden nodes + simulated annealing BM Architecture –a set of visible nodes: nodes can be accessed from outside.
AUTOWISARD: Unsupervised Modes for the WISARD Authors: Iuri Wickert Felipe M. G. França Computer Systems Engineering.
Robotica Lecture 3. 2 Robot Control Robot control is the mean by which the sensing and action of a robot are coordinated The infinitely many possible.
An Instructable Connectionist/Control Architecture: Using Rule-Based Instructions to Accomplish Connectionist Learning in a Human Time Scale Presented.
1 N -Queens via Relaxation Labeling Ilana Koreh ( ) Luba Rashkovsky ( )
CE Operating Systems Lecture 11 Windows – Object manager and process management.
AGI Architectures & Control Mechanisms. Realworld environment Anatomy of an AGI system Intellifest 2012 Sensors Actuators Data Processes Control.
Soft Computing Lecture 14 Clustering and model ART.
1 Learning Chapter 18 and Parts of Chapter 20 AI systems are complex and may have many parameters. It is impractical and often impossible to encode all.
Neural and Evolutionary Computing - Lecture 9 1 Evolutionary Neural Networks Design  Motivation  Evolutionary training  Evolutionary design of the architecture.
Precise and Approximate Representation of Numbers 1.The Cartesian-Lagrangian representation of numbers. 2.The homotopic representation of numbers 3.Loops.
Neural Networks Steven Le. Overview Introduction Architectures Learning Techniques Advantages Applications.
Introduction to Neural Networks and Example Applications in HCI Nick Gentile.
An Unsupervised Connectionist Model of Rule Emergence in Category Learning Rosemary Cowell & Robert French LEAD-CNRS, Dijon, France EC FP6 NEST Grant.
Human Abilities 2 How do people think? 1. Agenda Memory Cognitive Processes – Implications Recap 2.
Artificial Neural Networks Students: Albu Alexandru Deaconescu Ionu.
Face Image-Based Gender Recognition Using Complex-Valued Neural Network Instructor :Dr. Dong-Chul Kim Indrani Gorripati.
Dynamic Decision Making Laboratory Carnegie Mellon University 1 Social and Decision Sciences Department ACT-R models of training Cleotilde Gonzalez and.
C - IT Acumens. COMIT Acumens. COM. To demonstrate the use of Neural Networks in the field of Character and Pattern Recognition by simulating a neural.
Perceptrons Michael J. Watts
ACT-R 5.0 Architecture Christian Lebiere Human-Computer Interaction Institute Carnegie Mellon University
Multitasking Computational Neuroscience NSCI 492 Spring 2008.
Pattern Recognition. What is Pattern Recognition? Pattern recognition is a sub-topic of machine learning. PR is the science that concerns the description.
Deep Learning Overview Sources: workshop-tutorial-final.pdf
ACT-R 6.0 Software Updates Summer ‘09 – Summer ‘10 Dan Bothell Carnegie Mellon University
Supervised Learning – Network is presented with the input and the desired output. – Uses a set of inputs for which the desired outputs results / classes.
Does the brain compute confidence estimates about decisions?
National Taiwan Normal A System to Detect Complex Motion of Nearby Vehicles on Freeways C. Y. Fang Department of Information.
Symbolic Reasoning in Spiking Neurons: A Model of the Cortex/Basal Ganglia/Thalamus Loop Terrence C. Stewart Xuan Choo Chris Eliasmith Centre for Theoretical.
Big data classification using neural network
Neural Network Architecture Session 2
MXNet Internals Cyrus M. Vahid, Principal Solutions Architect,
Boltzmann Machine (BM) (§6.4)
Presentation transcript:

SAL: A Hybrid Cognitive Architecture Y. Vinokurov 1, C. Lebiere 1, D. Wyatte 2, S. Herd 2, R. O’Reilly 2 1. Carnegie Mellon University, 2. University of Colorado, Boulder

ACT-R: Overview An established, production rule-based cognitive architecture which implements a model of declarative memory. Created as a model of higher-level human cognition. Highly modular: ACT-R modules expose “buffers” to the central core of the system and the buffers can connect ACT-R to the outside world.

ACT-R: A Schematic View

ACT: Memory Theory Memory in ACT-R is stored in “chunks”; a chunk is just a data structure that contains some “slots” that are assigned values. Values can be any valid Lisp data structure including other chunks. When a retrieval request is made, the chunk with the highest activation is retrieved. Activation is calculated according to the formula: A = B + P + S + ε Where B is the base level activation, P is the activation due to partial matching, S is the spreading activation (uniformly 0 in our case), and ε is the noise.

ACT-R’s limitations ACT-R contains symbolic and subsymbolic components, but does not reach all the way down to the neural level. As a consequence, ACT-R doesn’t really have “eyes” or “hands” (motor module nonwithstanding). That makes it difficult to interact with the world in non-symbolic ways.

Enter Leabra Leabra (Local, Error-driven and Associative Biologically Realistic Algorithm) is a model of neural interaction developed by O’Reilly et. al. at the University of Colorado, Boulder. Emergent is an environment in which a Leabra model is realized. It can implement a self- contained, simulated 3D world. In particular, a model called LVis (Leabra Vision) implements a simulation of the human visual system.

The Leabra Vision Model

SAL: Synthesis of ACT-R and Leabra We combine ACT-R and Leabra by implementing an module that exposes a leabra-visual buffer to the ACT- R core. The module handles communication with Leabra using sockets; data is obtained from Leabra and commands are issued from ACT-R. Data taken from Leabra is transformed into chunks that are then made available in the leabra-visual buffer. The current integration only implements an interface to the vision model, but a neural motor module is in the works.

SAL Applications: Metacognition The Leabra neural network is trained to recognize 50 out of 100 object classes. The set of objects is thus partitioned into TRAIN and TEST subsets. The ACT-R model’s declarative memory is pre-loaded with examples of both TRAIN and TEST items. An ACT-R chunk obtained from Leabra observing an item contains parameters that measure the net activation of different layers of the network. ACT-R’s blending mechanism is used to determine whether the observed object belongs to the TRAIN or TEST class based on a recall cued on the aforementioned activations.

SAL Applications: Self-supervised Object Learning Goal: to ground symbol cognition in low-level perception. Three pre-training regimes were used to train the Leabra neural network: full training (recognition of all object classes), half-training (recognition of only 50 object classes), and no training (network weights are random). The set of objects presented to the model is a subset of object classes that the neural network was not trained to recognize, i.e. the TEST class. The chunk obtained from the observation contains a vector that represents the encoding of the visual stimulus in Leabra’s simulation of the inferotemporal cortex (IT) layer.

Self-supervised Learning, cont. When the integrated model observes a presented item, it tries to recall an association between the percept (i.e. the IT vector) and a label assigned to the item. If the model fails to recall an association (which will happen initially) it generates a label (in this case, simply an integer) to associate with the percept. The label is then used as feedback to the neural network, which adjusts its connection weights to increase the strength of association between the item and the label. During training, network weights converge to a stable representation of the IT feature vector for each object class. The complete model thus bootstraps from the initial feature set obtained from pre-training to learning, in a self-supervised fashion, to recognize object categories.

SAL Model Flow

Self-supervised Learning, cont. The pre-training regime with completely random network weights does not result in any learning at all. When the network is completely trained, the ACT-R model learns the labels almost perfectly, with the exception of shape-based confusions (globe/skull, toaster/dice). The half-trained model is the most interesting case.

Fully-trained model

Partially-Trained Model

Problems with the model The IT vector is a shape-based feature vector which does not capture orientation, size, or texture. We need another signal that will help us distinguish between objects. Like, say, a motor signal. We don’t have a neurally-based motor module so far, but we can do is mock up a symbolic motor module.

The Symbolic Motor Module The symbolic motor module is an extension of ACT-R that “acts” on the objects. It performs some symbolic operation on a presented object and returns either success or failure. The model remembers the results of actions just like it remembers the percept that it associates with a label. Recalls are then cued on both action results and visual percepts.

Confusion matrix showing the progress of self-supervised learning that combines the IT vector information with the symbolic motor module

Testing the model afterwards…

Future Work The next step of the SAL integration process is the creation of a neurally-based motor model in Leabra, which will interface with ACT-R via a buffer. The model is still in development. – But the model developer, Sergio Verduzco, was just indoctrinated trained at the ACT-R summer school We also aim to unify the metacognitive old/new recognition model with the self-supervised object learning model to improve performance. – Use all signal sources for maximal discrimination