Past Account

• Past AccountJul 26, 2021, 8:14 PM

  8 points

  on: What are some trig­gers that prompt you to do a Fermi es­ti­mate, or to pull up a spread­sheet and make a sim­ple/​rough quan­ti­ta­tive model?

  [Deleted]

• Past AccountJul 18, 2021, 8:30 PM

  3 points

  on: The Moun­taineer’s Fallacy

  [Deleted]

• Past AccountMay 31, 2021, 8:59 PM

  1 point

  in reply to: TurnTrout’s comment on: MDP mod­els are de­ter­mined by the agent ar­chi­tec­ture and the en­vi­ron­men­tal dynamics

  [Deleted]

• Past AccountMay 30, 2021, 1:21 PM

  LW: -1 AF: -3

  AF

  in reply to: TurnTrout’s comment on: MDP mod­els are de­ter­mined by the agent ar­chi­tec­ture and the en­vi­ron­men­tal dynamics

  [Deleted]

• Past AccountMay 29, 2021, 6:02 PM

  LW: 1 AF: -2

  AF

  in reply to: TurnTrout’s comment on: MDP mod­els are de­ter­mined by the agent ar­chi­tec­ture and the en­vi­ron­men­tal dynamics

  [Deleted]

• Past AccountMay 29, 2021, 12:12 AM

  LW: 1 AF: -2

  AF

  on: MDP mod­els are de­ter­mined by the agent ar­chi­tec­ture and the en­vi­ron­men­tal dynamics

  [Deleted]

• Past AccountMay 22, 2021, 11:15 PM

  1 point

  in reply to: Charlie Steiner’s comment on: The Vari­a­tional Char­ac­ter­i­za­tion of KL-Diver­gence, Er­ror Catas­tro­phes, and Generalization

  Because $f=\lambda \cdot \Delta L$. They are the same. Does that help?

• Past AccountMay 22, 2021, 2:03 PM

  6 points

  on: SGD’s Bias

  [Deleted]

• Past AccountMay 21, 2021, 2:23 PM

  1 point

  in reply to: Steveot’s comment on: The Vari­a­tional Char­ac­ter­i­za­tion of KL-Diver­gence, Er­ror Catas­tro­phes, and Generalization

  The term $\pi$ is meant to be a posterior distribution after seeing data. If you have a good prior you could take $\pi ={\pi }_0$. However, note $L\left(\pi \right)$ could be high. You want trade-off between the cost of updating the prior and the loss reduction.

  Example, say we have a neural network. Then our prior would be the initialization and the posterior would be the distribution of outputs from SGD.

  (Btw thanks for the correction)

• Past AccountMay 6, 2021, 2:11 PM

  4 points

  in reply to: ChristianKl’s comment on: Open and Wel­come Thread—May 2021

  [Deleted]

• Past AccountMay 6, 2021, 1:46 PM

  1 point

  in reply to: ChristianKl’s comment on: Open and Wel­come Thread—May 2021

  [Deleted]

• Past AccountApr 24, 2021, 3:48 PM

  1 point

  in reply to: interstice’s comment on: NTK/​GP Models of Neu­ral Nets Can’t Learn Features

  [Deleted]

• Past AccountApr 24, 2021, 1:28 AM

  1 point

  in reply to: interstice’s comment on: NTK/​GP Models of Neu­ral Nets Can’t Learn Features

  [Deleted]

• Past AccountApr 23, 2021, 8:39 PM

  1 point

  in reply to: interstice’s comment on: NTK/​GP Models of Neu­ral Nets Can’t Learn Features

  [Deleted]

• Past AccountApr 23, 2021, 4:36 PM

  4 points

  on: NTK/​GP Models of Neu­ral Nets Can’t Learn Features

  [Deleted]

• Past AccountMar 28, 2021, 10:00 PM

  1 point

  in reply to: johnswentworth’s comment on: What is the Differ­ence Between Cheer­ful Price and Shadow Price?

  Your example is interesting and clarifies exchange rates. However,

  The shadow price quantifies the opportunity cost, so if I’m paid my shadow price, then that’s just barely enough to cover my opportunity cost.

  This is an interpretive point I’d like to focus on. When you move a constraint, in this case with price, the underlying equilibrium of the optimization shifts. From this perspective your usage of the word ‘barely’ stops making sense to me. If you were to ‘overshoot’ you wouldn’t be optimal in the new optimization problem.

  At this point I understand that the cheerful price will be equivalent to or more than the shadow price. You want to be able to shift the equilibrium point and have slack left over. It just seems obvious, to me, that shadow price isn’t an exactly measurable thing in this context and so you’d naturally be led to make a confidence interval (belief) for it. Cheerful price is just the upper estimate on that. Hence, I’m surprised why this is being treated as a new /​ distinct concept.

• Past AccountMar 28, 2021, 9:50 PM

  4 points

  in reply to: Anon User’s comment on: What is the Differ­ence Between Cheer­ful Price and Shadow Price?

  I suppose this is the most correct answer. I’m not really updating very much though. From my perspective I’ll continue to see cheerful price as a psychological/​subjective reinvention of shadow price.

  Edit: It seems clear in this context, shadow price isn’t exactly measurable. Cheerful price is just the upper estimate on the shadow price.

• Past AccountMar 4, 2021, 5:48 AM

  1 point

  in reply to: Joar Skalse’s comment on: Why Neu­ral Net­works Gen­er­al­ise, and Why They Are (Kind of) Bayesian

  [Deleted]

• Past AccountFeb 28, 2021, 6:22 PM

  LW: 1 AF: 1

  AF

  in reply to: Joar Skalse’s comment on: Why Neu­ral Net­works Gen­er­al­ise, and Why They Are (Kind of) Bayesian

  [Deleted]

• Past AccountJan 26, 2021, 3:56 PM

  4 points

  in reply to: Eigil Rischel’s comment on: Rec­og­niz­ing Numbers

  [Deleted]