Bias and variance of the Bayesian-mean decoder

Authors: Arthur Prat-Carrabin, Michael Woodford

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental To assess the quality of the approximations presented above, we run simulations of an encodingdecoding process, and compare the approximated and true values of the bias and variance of the Bayesian-mean decoder, with different efficient encodings, and under different amounts of imprecision in the encoding.
Researcher Affiliation Academia Arthur Prat-Carrabin Department of Economics Columbia University New York, USA arthur.p@columbia.edu Michael Woodford Department of Economics Columbia University New York, USA mw2230@columbia.edu
Pseudocode No No pseudocode or algorithm blocks are present in the paper.
Open Source Code No Did you include the code, data, and instructions needed to reproduce the main experimental results (either in the supplemental material or as a URL)? [N/A]
Open Datasets No We study a case in which a normally-distributed stimulus is encoded through a normally-distributed representation. Specifically, the prior is a Gaussian distribution with mean m and standard deviation σ, i.e., x N(m, σ2).
Dataset Splits No The paper describes running simulations of a generative process and comparing analytical approximations to true values, but does not mention specific training, validation, or test dataset splits.
Hardware Specification No The paper mentions running simulations but does not provide any specific details about the hardware used, such as GPU/CPU models or other system specifications.
Software Dependencies No The paper does not provide specific software dependencies with version numbers used for the simulations, only general statements about numerical computation.
Experiment Setup Yes We study a case in which a normally-distributed stimulus is encoded through a normally-distributed representation. Specifically, the prior is a Gaussian distribution with mean m and standard deviation σ, i.e., x N(m, σ2). ... We run simulations in which the encoding noise, ν, spans a range of values: ν = 0.005, 0.01, 0.02, 0.05, and 0.1