CECN1 Projects

From Computational Cognitive Neuroscience Wiki

These simulation projects for the CECN textbook, implemented in Emergent serve as replacements for the original projects that were done in PDP++. They have completely self-contained documentation that opens up when the project is opened -- follow those instructions instead of the ones written in the textbook.

Contents 1 General Usage Tips 2 Chapter 2: Individual Neurons 3 Chapter 3: Networks of Neurons 4 Chapter 4: Hebbian Model Learning 5 Chapter 5: Error-Driven Task Learning 6 Chapter 6: Combined Model and Task Learning, and Other Mechanisms 7 Chapter 8: Perception and Attention 8 Chapter 9: Memory 9 Chapter 10: Language 10 Chapter 11: Higher Level Cognition

General Usage Tips

In addition to all the tips available at the Emergent website, the following are particularly relevant to these projects:

• To make the documentation instructions always available in a separate window: do Object/Edit Dialog in the menu just above the document text that shows up when you open the project. Alternatively, you can just always return to the document by clicking on the ProjectDocs tab at the top of the middle panel.

• If your screen is small, you may also want to make the document window "Always on Top": (e.g., in Linux, Emergent-logo-at-top-left/Advanced-> Keep Above Others) and also use the "Shade" feature: double-click the top control bar to toggle the window between "rolled up" and "rolled down" (also: Emergent-logo->Shade). (Caution: you cannot double click on the Emergent-logo itself!). Keeping the ProjectDocs window on the left over the tree browser panel frame seems to work best since you will generally be watching and working with the right two frames the most.

• Projects for chapters 2-5 require the 4.11 version of Emergent, while those from chapter 6 onward require 4.13 (released 2/21/08), and 4.13 is recommended for all users.

Chapter 2: Individual Neurons

• CECN1 Units (units.proj) -- Individual Neuron (Unit)
• CECN1 Detector (detector.proj) -- The Unit as a Detector
• CECN1 Self Reg (self_reg.proj) -- Self regulation through accommodation and hysteresis (optional)

Chapter 3: Networks of Neurons

• CECN1 Transform (transform.proj) -- Transformations produced by feedforward neural processing
• CECN1 Localist vs Distributed (loc_dist.proj) -- Localist vs Distributed representations
• CECN1 Bidirectional Transformations (bidir_xform.proj) -- Bidirectional (top-down and bottom-up) processing and transformations
• CECN1 Pattern Completion (pat_complete.proj) -- Pattern completion
• CECN1 Top-down Amplification (amp_top_down.proj) -- Top-down amplification of bottom-up signals
• CECN1 Distributed Top-down Amplification (amp_top_down_dist.proj) -- Top-down amplification in distributed network
• CECN1 Inhibition (inhib.proj) -- Inhibitory interactions
• CECN1 Inhibition Digits (inhib_digits.proj) -- Digits model revisited with inhibition
• CECN1 Cats and Dogs (cats_and_dogs.proj) -- Constraint satisfaction in the Cats and Dogs model
• CECN1 Necker Cube (necker_cube.proj) -- Constraint satisfaction and the role of noise and accommodation in the Necker Cube model

Chapter 4: Hebbian Model Learning

• CECN1 Hebbian Correlation (hebb_correl.proj) -- Hebbian correlational learning
• CECN1 Self Organizing (self_org.proj) -- Self organizing learning using Hebbian learning and inhibitory competition

Chapter 5: Error-Driven Task Learning

• CECN1 Pattern Associator (pat_assoc.proj) -- Simple input/output pattern association task learning
• CECN1 Generec (generec.proj) -- Full error-driven learning using the Generalized Recirculation algorithm

Chapter 6: Combined Model and Task Learning, and Other Mechanisms

• CECN1 Model And Task (model_and_task.proj) -- Combined model and task learning
• CECN1 Family Trees (family_trees.proj) -- Learning in a deep (multi-hidden-layer) network
• CECN1 FSA (fsa.proj) -- Finite State Automaton test of Simple Recurrent Networks
• CECN1 Reinforcement Learning (rl_cond.proj) -- Simple Pavlovian Conditioning using Temporal Differences Learning

Chapter 8: Perception and Attention

• CECN1 V1Rf (v1rf.proj) -- V1 receptive fields from Hebbian learning
• CECN1 Objrec (objrec.proj) -- invariant object recognition
• CECN1 AttnSimple (attn_simple.proj) -- simple attention model
• CECN1 Objrec Multiobj (objrec_multiobj.proj) -- Object recognition with multiple objects: NOTE Not converting from PDP++ until later (click for more info)

Chapter 9: Memory

• CECN1 Wt Priming (wt_priming.proj) -- Weight-based (long-term) priming
• CECN1 AB-AC List Learning (ab_ac_interference.proj) -- Paired associate learning and catastrophic interference
• CECN1 Hippocampus (hip.proj) -- Hippocampus model and overcoming interference
• CECN1 Act Priming (act_priming.proj) -- Activation-based (short-term) priming
• CECN1 Active Maintenance (act_maint.proj) -- Active maintenance in simple prefrontal cortex (PFC)
• CECN1 PFC Maint Updt (pfc_maint_updt.proj) -- Updating and Maintenance in more complex PFC model
• CECN1 A Not B (a_not_b.proj) -- Development of interacting memory systems and the A-not-B task

Chapter 10: Language

• CECN1 Dyslexia (dyslex.proj) -- Normal and disordered reading and the distributed lexicon
• CECN1 Spelling to Sound (ss.proj) -- Orthography to Phonology mapping and regularity, frequency effects
• CECN1 Past Tense (pt.proj) -- Overregularization of the English past tense inflection
• CECN1 Semantics (sem.proj) -- Semantic Representations from World Co-occurrences and Hebbian Learning
• CECN1 Sentence Gestalt (sg.proj) -- The Sentence Gestalt model

Chapter 11: Higher Level Cognition

• CECN1 Stroop (stroop.proj) -- The Stroop effect and PFC top-down biasing
• CECN1 ID_ED (id_ed.proj) -- The intradimensional/extradimensional task