Week 6: Writing Assignment

Answer ONE of the following questions based on this week's material (one-page MAX, PDF format, email to ajyu@ucsd.edu with Subject "Cogs 160: HW 6", due next Friday). State clearly which question you are responding to. You may also answer up to two bolded questions on the wiki for week 5, as extra credit (10% each).

1. Summarize the evidence that musically trained subjects have enhanced subcortical representations of pitch, timbre, and timing, compared to non-musicans.
2. Is the enhanced representation of pitch, timbre, and timing in musicians, compared to non-musicians, the resut of a self-selection process (those better at these things are the ones that become musicians), or the causal consequences of musical training? Provide experimental evidence to support your statement.
3. What is the efficient coding hypothesis? What does it mean in the context of tuning curve for the grasshopper auditory receptor? Discuss this for the two cases of with and without noise in neural response. (See Figure 1 of Machens et al, 2005)
4. Is the fact that r-OSE's collapse onto a single solution when input solution when input space is transformed by "clipping" surprising? Explaining why or why not.
5. Use Fig. 4 of Machens et al (2005) to explain why OSE should prefer maximal output entropy while minimizing noise entropy. Why do the two trade off?
6. How can Machens et al's claim that "auditory neural code is optimized for behavioral relevant stimuli" be further substantiated? Describe what kind of new data they could collect, and/or analysis they could perform.

Last modified: Mon May 9 01:51:41 EDT 2011