Learning a neural response metric for retinal prosthesis

Authors: Nishal P Shah, Sasidhar Madugula, EJ Chichilnisky, Yoram Singer, Jonathon Shlens

ICLR 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Using data from electrical stimulation experiments, we demonstrate that the learned metric could produce improvements in the performance of a retinal prosthesis.3.1 EXPERIMENTAL SETUP Spiking responses from hundreds of retinal ganglion cells (RGCs) in primate retina were recorded using a 512 electrode array system (Litke et al., 2004; Frechette et al., 2005).
Researcher Affiliation Collaboration 1Stanford University 2Google Brain 3Princeton University
Pseudocode No The paper describes the network architecture in Figure 6 and Table 1 but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code (no specific repository link, explicit code release statement, or code in supplementary materials).
Open Datasets No Spiking responses from hundreds of retinal ganglion cells (RGCs) in primate retina were recorded using a 512 electrode array system (Litke et al., 2004; Frechette et al., 2005).
Dataset Splits No The responses were partitioned into training (first 8 seconds) and testing (last 2 seconds) of each trial.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types, or memory) used for running its experiments.
Software Dependencies No Optimizer Adam (Kingma & Ba, 2014) (α = 0.01, β1 = 0.9, β2 = 0.999) Parameter updates 20,000 Batch size 100 Weight initialization Xavier initialization (Glorot & Bengio, 2010) (The paper mentions optimizers and initialization methods, but does not provide specific software dependencies with version numbers for reproducibility, such as Python or library versions.)
Experiment Setup Yes Optimizer Adam (Kingma & Ba, 2014) (α = 0.01, β1 = 0.9, β2 = 0.999) Parameter updates 20,000 Batch size 100 Weight initialization Xavier initialization (Glorot & Bengio, 2010)