Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Safe Exploration in Dose Finding Clinical Trials with Heterogeneous Participants
Authors: Isabel Chien, Wessel P Bruinsma, Javier Gonzalez, Richard E. Turner
ICML 2024 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate SAFE-T on a thorough set of synthetic scenarios (Sec. 5.1), differing subgroup distributions (Sec. 5.2), and sample size variations (Sec. 5.3). In Sec. 5.4, we apply SAFE-T to a new adaptive setting, demonstrating possible future extensions. |
| Researcher Affiliation | Collaboration | 1University of Cambridge, Cambridge, UK 2Microsoft Research AI for Science 3Microsoft Research. |
| Pseudocode | Yes | We first discuss important components of SAFE-T and then detail the algorithm in Section 3.2, with pseudocode in Algorithm 1. |
| Open Source Code | No | The paper does not provide any explicit statements about open-source code availability or links to a code repository for the methodology described. |
| Open Datasets | No | The paper uses 'synthetic scenarios' which are constructed by the authors based on literature, but no concrete access information (link, DOI, repository, or formal citation to a public dataset) is provided for these scenarios. |
| Dataset Splits | No | The paper does not specify explicit training, validation, or test dataset splits (e.g., percentages or sample counts) for the synthetic scenarios used in the experiments. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory, or cloud instance types) used for running the experiments. |
| Software Dependencies | No | The paper mentions software like 'Py MC' and 'GPy Torch' for implementation but does not provide specific version numbers for these or other software dependencies. |
| Experiment Setup | Yes | We define νT = 0.2 for the safety constraint and νE = 0.2 (needed when UCB efficacy optimization is used)... Both the toxicity and efficacy GPs use constant mean functions (we set mean = 0.3 for toxicity and mean = 0.1 for efficacy) and the stationary radial basis function kernel (RBF kernel) as the covariance function (we set length scale = 4 for toxicity and length scale = 2 for efficacy). We also set the matrix A, with Q rows and S columns, which is composed of the coefficients a(i) s of the LMC model to 1.0 0 0.2 0.2 0 1.0. |