Skill UI

This guide details the methodology for packaging scientific artifacts (code, solvers, data, logs) for machine learning submissions. It focuses on creating double-blind, self-contained ZIP archives that allow reviewers to independently and quickly verify complex probabilistic claims, even under strict time constraints. It covers best practices for anonymity, structured README generation, and handling data/licensing.

This guide provides a strategic framework for drafting author responses to challenging peer reviews in academic conferences. It teaches authors how to structure replies by addressing four key axes (Correctness, Novelty, Backing, Clarity), ensuring technical precision, maintaining anonymity, and framing commitments that survive the subsequent Area Chair (AC) discussion phase.

A comprehensive guide for authors submitting to UAI (Uncertainty in Artificial Intelligence). It details how to position research by ensuring coverage across diverse academic communities, including Machine Learning, Statistics, Causal Inference, and foundational mathematics. Focuses on establishing the technical 'delta' rather than merely listing citations, helping authors avoid common review failures.

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.

A comprehensive guide for authors submitting to venues focused on probabilistic AI, including UAI. It emphasizes achieving maximum formal rigor by enforcing strict notational discipline, clearly separating conditioning from intervention, explicitly stating assumptions, structuring the main body for maximum impact, and calibrating claims to match empirical evidence. Essential for papers involving graphical models, causality, and Bayesian methods.

A comprehensive guide for structuring the 'Related Work' section for top-tier HCI conferences (e.g., UIST). This document emphasizes that the section must function as a novelty proof, not a literature survey. It teaches authors to organize content by capability delta, utilize detailed comparison tables, and trace three critical lineages: technique, toolkit, and commercial practice, ensuring the work's unique contribution is clearly established.

This guide provides a rigorous methodology for achieving scientific reproducibility in technical and user experience papers (e.g., UIST). It details three essential 'ledgers': the Build Ledger (pinning software, hardware, and environment dependencies), the Measurement Ledger (defining precise protocols for metrics like latency and accuracy), and the Study Ledger (documenting human evaluation tasks). Following these protocols ensures that external researchers can accurately rebuild the artifact and validate reported results.

A comprehensive guide to navigating the UIST (User Interface Software and Technology) academic review pipeline. It details every stage, from initial submission and expert review to the rebuttal phase and final committee decisions. The guide teaches authors how to critically analyze reviewer feedback, identify evidence gaps, strengthen novelty claims, and strategically plan for conditional acceptance or resubmission.

A comprehensive guide detailing the standards for scientific reproducibility in visualization research, particularly for IEEE VIS submissions. It outlines the Graphics Replicability Stamp Initiative (GRSI) requirements, covering how independent volunteers must reproduce results using provided code, data, and environments (e.g., Docker containers). Learn how to package your work—including determinism aids, mapping tables, and licensed data—to pass rigorous, clean-machine testing.

A comprehensive guide for authors submitting to top-tier visualization venues (like IEEE VIS/TVCG). It teaches authors how to write the 'Related Work' section by adopting a 'delta-first' contrast approach, demonstrating novelty by clearly positioning the new work against prior art across multiple adjacent domains, including HCI, Graphics, and other regional visualization venues.

This guide details best practices for achieving high standards of reproducibility in visual analytics papers, focusing on open materials, verifiable code, and rigorous documentation. It covers mapping figures to data sources, preregistering studies, documenting data provenance, and defining the level of reproducibility achieved (e.g., turnkey, scripted).

This guide outlines the rigorous process for authors responding to a "revise" verdict at top-tier academic conferences. It teaches authors how to treat reviewer feedback as a formal contract, manage a tight revision timeline, structure necessary experiments, and write a highly traceable change document that explicitly addresses every required point.