The Value of Information When Deciding What to Learn

Authors: Dilip Arumugam, Benjamin Van Roy

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

Reproducibility Variable Result LLM Response
Research Type Experimental We conclude in Section 4 with a corroborating set of computational experiments that demonstrate the efficacy of our approach.
Researcher Affiliation Academia Dilip Arumugam Department of Computer Science Stanford University dilip@cs.stanford.edu Benjamin Van Roy Department of Electrical Engineering Department of Management Science & Engineering Stanford University bvr@stanford.edu
Pseudocode No The paper describes algorithms in text but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper mentions using a third-party Python package 'dit' and provides a link to its GitHub repository (https://github.com/dit/dit) in a footnote. However, it does not state that the code for the methodology described in this paper (BLAIDS implementation) is open-source or provide a link to its own code.
Open Datasets No The paper refers to
Dataset Splits No The paper does not provide specific details on dataset splits (e.g., percentages, sample counts for training, validation, or testing sets). It mentions averaging results over "10 random trials" but this refers to experiment repetitions, not data splits.
Hardware Specification No The paper states, "run experiments via Google Colab with a GPU." This mention of "a GPU" is too general and does not specify a particular model or detailed hardware specifications.
Software Dependencies No The paper mentions using 'Adam' and the 'dit' Python package (referencing [James et al., 2018]), but it does not provide specific version numbers for these software components in the main text required for reproducible setup.
Experiment Setup Yes We found the use of a linear hypermodel [Dwaracherla et al., 2020] to be sufficient when optimized via Adam [Kingma and Ba, 2014] (learning rate of 0.001) with a noise variance of 0.1, prior variance of 1.0, and batch size of 1024.