Skill UI

This comprehensive skill guides authors through the specific submission requirements and editorial culture of the Journal of Machine Learning Research (JMLR). It helps authors determine if their ML method, theory paper, or dataset is appropriate for JMLR's high standards—emphasizing rigor, complete proofs, open code, and lasting scientific contribution over novelty signaling. Use this when targeting JMLR or comparing it against other top-tier ML venues.

A comprehensive guide for structuring the empirical section of high-level machine learning papers (e.g., KDD). It details the 'four-axis evidence plan'—Quality, Scale, Efficiency, and Mechanism—and emphasizes critical data hygiene practices. Learn how to prevent temporal leakage, structure ablations in a matrix format, and design post-launch measurements to ensure academic rigor and reproducibility.

A comprehensive guide for authors applying machine learning techniques to healthcare. It helps authors frame their research to meet the high standards of top ML4H conferences, covering biomedical data leakage, confounding effects, required evidence standards, and optimal submission strategies.

A comprehensive guide for authors packaging their accepted MLSys research artifacts. It details strategies for achieving Availability, Functional, and Reproducible badges by correctly structuring the submission, handling hardware dependencies, and writing the detailed Artifact Appendix. Essential for ensuring scientific claims are independently verifiable.

Provides a comprehensive, time-sensitive playbook for authors responding to reviews (rebuttals) at MLSys. It details how to triage objections, prioritize critical missing measurements, address systems objections regarding workload representativeness and baseline tuning, and structure anonymous, data-driven replies within a compressed rebuttal window.

This guide details the critical steps for authors preparing their accepted paper for final publication in MLSys proceedings. It covers de-anonymization, restoring company and repository identity, reconciling promised rebuttals, and ensuring metadata accuracy for long-term findability. Essential for managing the complex post-acceptance timeline.

This comprehensive guide instructs researchers on how to design rigorous and credible evaluations for Machine Learning systems. It covers critical areas such as workload selection, baseline comparison discipline, reporting four key coordinates (throughput, latency tails, memory, cost), structuring ablations, and generating scaling evidence. It ensures measurements meet academic benchmarking standards.

A comprehensive guide for authors to position their ML systems research by mapping the vast, scattered literature across multiple venues (OSDI, SOSP, NeurIPS, open-source projects). It details how to audit novelty, handle fast-moving advancements, and write a defensible delta statement against both academic and industrial prior work.

Provides comprehensive guidelines for achieving high reproducibility in Machine Learning Systems (MLSys) performance claims. It details how to rigorously control sources of non-determinism (e.g., thermal noise, random seeds) and pin the entire technical stack—from hardware and drivers to frameworks and serving policies—to ensure results can be validated by third parties.

A comprehensive guide to modeling the specific peer review cycle of MLSys submissions. It details the OpenReview workflow, the challenges posed by the mixed ML and systems reviewer pool, and provides strategic advice on addressing concerns related to workload representativeness, baseline strength, and evidence scope matching for high-stakes technical papers.

A comprehensive pre-deadline audit checklist for submitting papers to the MLSys conference. It guides authors through track-specific requirements (Research vs. Industrial), formatting compliance (10-page limit, two-column layout), maintaining double-blind anonymity, and correctly handling appendices and citations to ensure readiness before the submission deadline.

This guide provides critical structuring advice for authors submitting ML Systems papers, focusing on the strategic division of content between the mandatory 10-page body and the unlimited supplementary appendix. It outlines which material belongs in the main text (claims that must be believed) versus what belongs in the appendix (material that must be verifiable evidence, such as full hyperparameter grids or deep ablations).