This is our wonderfully ambitious schedule.
We'll attempt to keep with it, but it is subject to modification.

Date Topic Readings
(* = to be read by this class.
All others are reference readings
)
Notices &
Reading Questions
1/6/15 Admin &
Selection of articles
[Lisa & Mark]

Introductory message board discussion points due
1/8/15 Underlying principles: Levels
[Lawrence]

* (1) Griffiths et al. in press
(2) Marr 1982, Ch.1: only pp. 24-29
(3) Bonawitz et al. 2014
Message board discussion points due
  • What are the three levels of analysis?
  • How do they relate to rational models vs. process models?
1/13/15 Underlying principles:
Bayesian inference
[Colin]

* (1) Griffiths et al. 2012
* (2) Bowers & Davis 2012
(3) Bowers & Davis 2012 original article
(4) Bever & Poeppel 2010
(5) Griffiths et al. 2010
Message board discussion points due
  • What are some of the main myths about the Bayesian approach?
  • What about its strengths?
  • How does using Bayesian methods as a statistical tool differ from using it as a theory of cognition?
1/15/15 Underlying principles:
Bayesian inference
[Garren]

* (1) Perfors 2012
(2) Griffiths et al. 2008
(3) Griffiths et al. 2012

Message board discussion points due
  • Why should every cognitive scientist who believes in the Bayesian approach be called a nativist? (What does it mean to be a nativist?)
  • If the hypothesis space is specified, what does learning consist of?
  • Can the hypothesis space itself change?
1/20/15 Underlying principles:
Deep Belief Nets
[Joseph]
(pdf)
* (1) Hinton 2006
(2) Saxe et al. 2013
(3) McClelland et al. 2010
(4) Szegedy et al. 2013
(5) Videos of Deep Belief Nets in action: Generating digits, Generating walks
(6) Another demo: Automatically generating text for an image + Kiros et al. 2014, which describes models combining image and text learning

Message board discussion points due
1/22/15 Language:
Acquisition
[Ereka]

* (1) Pearl & Goldwater in press
(2) Perfors 2014a
(3) Perfors 2014b
(4) Tenenbaum & Griffiths 2001
Message board discussion points due
  • How do hard and soft constraints correspond to the hypothesis space of the learner?
  • How do these relate to the Bayesian learner's prior?
  • How does optimality play into inference?
  • How does Bayesian learning relate to indirect negative evidence and the poverty of the stimulus?
1/27/15 Cognitive development
[Pele]

* (1) Gopnik & Bonawitz 2014
(2) Perfors et al. 2011
(3) Piantadosi et al. 2012
(4) Kemp & Tenenbaum 2009
Message board discussion points due
  • How do probabilistic models mediate between the extreme versions of the nativist and empiricist approaches?
  • How does a framework differ from a model?
  • What are the aspects of a model that must be made explicit?
  • How are transitions in cognitive development represented?
  • What is the blessing of abstraction?
1/29/15 Causality
[BreAnna]

* (1) Gopnik & Wellman in press
(2) Xu et al. 2009
(3) Goodman et al. 2011
(4) Kemp et al. 2010
Message board discussion points due
  • What is rational constructivism?
  • What are the main aspects of the theory theory?
  • What is the inverse problem?
  • Why is the hierarchical aspect of probabilistic models so important?
  • How does it relate to the blessings of abstraction?
2/3/15 Language: Use
[Prutha]

* (1) Goodman & Lassiter in press
(2) Bergen et al. 2014
(3) Myslin & Levy in press
(4) Bergen et al. 2012
(5) Goodman & Stuhlmuller 2013
(6) Frank & Goodman 2012 + supplementary material
(7) Kao et al. 2014
(8) Probabilistic Models of Cognition:
Overview
Intro to lambda calculus
Inference about inference
Non-parametric models

Message board discussion points due
  • How does lambda calculus relate to the conceptual lexicon?
  • What are possible worlds and why do they matter?
  • What is the difference between pragmatic interpretation and literal meaning?
  • How does the pragmatic interpretation relate to a rational model of language understanding?
2/5/15 Theory of Mind: Social goals [Galia]

* (1) Baker et al. 2008
Message board discussion points due
2/10/15 Theory of Mind: Decisions
[Prachi]

* (1) Jern & Kemp 2011
Message board discussion points due
2/12/15 Decisions: Heuristics
[Irina]

* (1) Gigerenzer & Brighton 2009
(2) Lieder et al. 2014
(3) Gigerenzer 2008

Message board discussion points due
2/17/15 Decisions: Predictions
[Percy]

* (1) Griffiths & Tenenbaum 2006

Message board discussion points due
2/19/15 NO CLASS



Be working on the readings for next time.
2/24/15 Language: Comprehension
[K.J.]

* (1) Levy 2013
(2) Levy 2011
(3) Fossum & Levy 2012
Message board discussion points due
  • How do the different memory-based approaches work?
  • How do the expectation-based approaches work?
  • When they come into conflict, which one(s) better explain(s) the observed data?
  • How do expectation-based approaches relate to generative models (rational comprehension)?
2/26/15 Memory
[Stephen]

Special location:
SBSG 1321


* (1) Hemmer & Steyvers 2009
(2) Steyvers & Hemmer 2012


Message board discussion points due
3/3/15 Collective intelligence
[Irina]

* (1) Steyvers & Miller in press
(2) Yi et al. 2012


Message board discussion points due
3/5/15 Expertise
[Stephen]

* (1) Lee et al. 2014
(2) Lee et al. 2012


Message board discussion points due
3/10/15 Connecting to neuroscience
[Garren]

* (1) Turner et al. 2013

Message board discussion points due
3/12/15 Peer review
[Everyone]

  • Upload your draft to the dropbox folder under SharedStudentFiles labeled "Drafts of Writing Assignments" by 3:30pm.

  • Bring your laptops to class so you can create text files with comments on other people's drafts. These will then be uploaded to the dropbox folder under SharedStudentFiles labeled "Peer Review Comments".
    Assigned reviews are here.
    Comments due 3/13/15 @ 3:30pm.
3/17/15 Final presentations
[Everyone]

  • Final presentation slides should be uploaded to the dropbox folder under SharedStudentFiles labeled "Final Presentation Slide" by 4pm.

  • Final writing assignment should be uploaded to the dropbox folder under AssignmentSubmission labeled "Final Writing Assignment" by 6pm.