Skill UI

A comprehensive guide for authors submitting to machine learning conferences like NeurIPS. It details the rigorous standards for data splitting, hyperparameter reporting, random-seed protocols, and compute transparency required for a main-track paper. It also helps authors strategically decide between the main track and the MLRC/TMLR routes based on their core contribution (replication vs. novelty).

A comprehensive guide for computer graphics researchers detailing how to achieve result-level reproducibility, especially for major conferences like SIGGRAPH. It emphasizes pinning all provenance—including assets, random seeds, hardware stacks, and hyperparameters—and methodologies for handling non-determinism in floating-point math and stochastic simulations, ensuring that the reported figures and timings can be regenerated by any reader.

A comprehensive guide for ensuring the reproducibility of Information Retrieval (IR) research, essential for SIGIR submissions. It details best practices for pinning data collection versions, index settings, neural model hyperparameters, and managing non-determinism. It helps researchers document artifacts—not just promises—to validate reported gains.

This guide details advanced best practices for achieving scientific reproducibility in probabilistic modeling and deep learning research. It mandates rigorous disclosure of seeds, hyperparameters, diagnostic traces (e.g., R-hat, ELBO), and compute environments to ensure that all claims made in academic papers are verifiably backed by traceable evidence. It elevates reproducibility from a simple code run to a comprehensive audit.