...consider a subject trained to choose between objects A and B, where A is $1,000,000 worth of goods and B is $100,000 worth of goods… A system that represented only the relative expected subjective value of A and B would represent SV(A) > SV(B). Next, consider training the same subject to choose between C and D, where C is $1,000 worth of goods and D is $100 worth of goods. Such a system would represent SV(C) > SV(D). What happens when we ask a chooser to select between B and C? For a chooser who represents only relative expected subjective value, the choice should be C: she should pick $1,000 worth of goods over $100,000 worth of goods because it has a higher learned relative expected subjective value. In order for our chooser to… construct transitive preferences across choice sets (and to obey the continuity axiom)… it is required that somewhere in the brain she represent the absolute subjective values of her choices.
This seems to be refuting a strawman. What does “representing only relative subjective value” even mean?
I’d say:
Transitivity of preference is obviously good.
I’ve noticed that I tend to compare experiences to memories of similar experiences. Likely I’d be comparing memory against memory.
Refuting a strawman? Whom is this passage (by Glimcher, not Tobler) supposed to be refuting? I’m not sure anyone would disagree with this.
‘Representing only relative subjective value’ means that the system would only encode subjective value (utility), not objective value. Representation in this case would occur via neuron firing rates.
Strawman position. They made up the position then refuted it.
The possibility being argued against is, as I understand it:
We remember the value of everything we’ve decided between as follows: every time we compare it in a choice against alternatives, we remember how strongly it beat the alternatives. And that’s all we remember. When we next evaluate that thing, if we have a memory of it, that’s all we use—that coded value of how much we (dis)preferred it against some alternatives (which we’ve forgotten).
This is very bad. It’s not even complete. Where does the first preference come from? Why even discuss such a thing? To fix the above, we can say that the preference is a weighted combination of choice-amnesiac-desirability (as if we have no memories of ever having (dis)preferred the thing to other things in the past), and all past choices pro/con that thing (without reference to the competitors’ values). This is now well defined, and perhaps worth ruling out by experiment.
Lukeprog quoting Tobler:
This seems to be refuting a strawman. What does “representing only relative subjective value” even mean?
I’d say:
Transitivity of preference is obviously good.
I’ve noticed that I tend to compare experiences to memories of similar experiences. Likely I’d be comparing memory against memory.
Refuting a strawman? Whom is this passage (by Glimcher, not Tobler) supposed to be refuting? I’m not sure anyone would disagree with this.
‘Representing only relative subjective value’ means that the system would only encode subjective value (utility), not objective value. Representation in this case would occur via neuron firing rates.
Strawman position. They made up the position then refuted it.
The possibility being argued against is, as I understand it:
We remember the value of everything we’ve decided between as follows: every time we compare it in a choice against alternatives, we remember how strongly it beat the alternatives. And that’s all we remember. When we next evaluate that thing, if we have a memory of it, that’s all we use—that coded value of how much we (dis)preferred it against some alternatives (which we’ve forgotten).
This is very bad. It’s not even complete. Where does the first preference come from? Why even discuss such a thing? To fix the above, we can say that the preference is a weighted combination of choice-amnesiac-desirability (as if we have no memories of ever having (dis)preferred the thing to other things in the past), and all past choices pro/con that thing (without reference to the competitors’ values). This is now well defined, and perhaps worth ruling out by experiment.
I too am having trouble understanding what the “relative subjective value” hypothesis is supposed to be.