UIST Reproducibility

•
1.1k
•

Standardizes artifact reproducibility, measurement protocols, and hardware specifications for academic UIST system papers.

UIST Reproducibility guides researchers and interface engineers in documenting systems for full replicability in ACM UIST submissions and human-computer interaction (HCI) publications. It structures documentation across three distinct ledgers—the build ledger (software, hardware BOM, environmental assumptions, and tuning constants), the measurement ledger (deterministic evaluation harnesses, raw timestamped logs, and clear latency/accuracy boundaries), and the study ledger (experimental protocols, counterbalancing, and IRB-compliant availability statements). This skill is particularly useful for pre-submission audits, preparing supplementary materials, creating one-page REPRO.md manifests, and crafting honest availability statements that preempt reviewer skepticism.

Key Features

01Measurement protocol pinning for latency, recognition accuracy, tracking error, and throughput
021,090 GitHub stars
03Study ledger structuring for user study instruments, stimuli, and IRB-compliant data statements
04Automated one-page REPRO.md manifest generation for supplementary submission packages
05Build ledger tracking covering hardware BOMs, firmware, software lockfiles, and magic numbers
06Integrated ML reporting guidelines for systems with embedded learned models and third-party APIs

Use Cases

01Drafting transparent and credible data and code availability statements for academic reviews
02Auditing a draft UIST or HCI paper to identify unstated parameters before submission
03Creating a complete supplementary package with automated rerun scripts and REPRO.md manifests