Modeling Joseph Glavan Joseph Houpt Wright State University August 7 2016 Purpose The TimeBased ResourceSharing TBRS model is a mostly verbal model of working memory Formalize into an endtoend computational model using ACTR ID: 1038522
Download Presentation The PPT/PDF document "Exploring the Time-Based Resource-Sharin..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. Exploring the Time-Based Resource-Sharing Model through Computational ModelingJoseph GlavanJoseph HouptWright State UniversityAugust 7, 2016
2. PurposeThe Time-Based Resource-Sharing (TBRS) model is a mostly verbal model of working memoryFormalize into an end-to-end computational model (using ACT-R)Highlight ancillary assumptionsIdentify experiments to test these assumptions and further constrain the theoryBarrouillet, Bernardin & Camos (2004); Oberauer & Lewandowsky (2011)2
3. TBRS ModelAttention is needed for maintenance and processingCentral bottleneckTemporal decayRapid switching between maintenance and processingBarrouillet, Bernardin & Camos (2004)3
4. Cognitive Load4k = raw WM capacity (unique to individual, task, etc.)Barrouillet, Portrat & Camos (2011)
5. Y = -8.33 * X + 8.13R2 = .9855Barrouillet, Portrat & Camos (2011)
6. The TaskBarrouillet et al. (2007) – Experiment 3Designed to compare effects of different sources of CLSix between-subjects conditions2 CL sources (retrievals vs response selection)3 paces of distractorsTrained on 96 judgment trials and 1- and 2-letter setsRecall7b*750 ms1500 msdelay500 msdelay500 ms6400 msNext target6
7. ACT-R77
8. Model Behavior88
9. Model BehaviorList order preserved throughEpisodic similarityTemporal inhibition99
10. Model Results10
11. 1111
12. 1212
13. DiscussionGood qualitative fit, poor quantitative fitList representation underspecified in TBRSSerial position effects/errorsSelf-generated reinforcement inflates CLNon-attentional mechanisms (articulatory rehearsal)13Camos & Barrouillet (2014)
14. Future WorkExplore full range of CL within-subjectsCompare temporal decay and representational interference using ACT-RGeneral attentional refreshing routine for ACT-RCan use the model to measure CL in applied settings14
15. Questions?
16. ReferencesAnderson, J. R., Bothell, D., Byrne, M. D., Douglass, S., Lebiere, C., & Qin, Y. (2004). An integrated theory of the mind. Psychological review, 111(4), 1036.Barrouillet, P., Bernardin, S., & Camos, V. (2004). Time constraints and resource sharing in adults' working memory spans. Journal of Experimental Psychology: General, 133(1), 83.Barrouillet, P., Bernardin, S., Portrat, S., Vergauwe, E., & Camos, V. (2007). Time and cognitive load in working memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(3), 570.Barrouillet, P., Portrat, S., Camos, V. (2011). On the law relating processing to storage in working memory. Psychological Review, American Psychological Association, 118 (2), pp.175-92.Camos, V., & Barrouillet, P. (2014). Attentional and non-attentional systems in the maintenance of verbal information in working memory: the executive and phonological loops. Frontiers in Human Neuroscience, 8, 900. doi:10.3389/fnhum.2014.00900Oberauer, K., & Lewandowsky, S. (2011). Modeling working memory: a computational implementation of the time-based resource-sharing theory. Psychonomic Bulletin & Review, 18(1), 10-45.
17. Perception Subroutine17
18. New-List Subroutine18
19. Target-Related Subroutine19Example Target ChunkChunk Name:TARGET1-0Chunk Type:targetChunk Slots (features)string“c”parentstartlist“1000”episode1.941
20. 20Distractor-Related Subroutine:Parity condition
21. 21Distractor-Related Subroutine:Spatial-location condition
22. Procedural Learning Mechanisms22Utility learningResponsible for learning the correct responseUi(n) = Ui(n-1) + α [Ri – Ui(n-1)]Production compilationResponsible for learning to respond fasterBypass retrievals
23. Recall Subroutine23
24. Maintenance Subroutine24
25. Retrieval Equations25
26. Stable LTM Assumption26
27. FixedFreeLearning rate = .2Utility noise = 1Base-level decay = .5Inhibition-scale = 1Association scale = 1Activation noise = .3Retrieval threshold = 0Reward ( R )Inhibition-decay (γ)Base-level constant (β) Episodic selectivity (η)Latency-exponent ( f )Latency-factor ( F )Model Parameters27
28. Predictions of Span28Barrouillet et al., 2007; Barrouillet, Portrat & Camos, 2011
29. Predictions of Span29Barrouillet et al., 2007; Barrouillet, Portrat & Camos, 2011