MLSys Skills
A 12-skill depth pack for MLSys (Conference on Machine Learning and Systems) submissions: venue fit at the ML-systems intersection, OpenReview submission checks, the short author-response window, camera-ready, badge-driven artifact evaluation, performance-measurement reproducibility, appendix packaging, review process, systems writing style, related work, benchmark-rigor experiments, and cycle wor
Skills in this plugin
12- ▌ Mlsys Workflow · brycewang-stanfordUse when planning an MLSys submission cycle end to end, from venue-fit and track choice through evaluation freeze, the October deadline, the four-day January response, notification, camera-ready, the March artifact-evaluation submission, and the May conference, with backward planning built around scarce GPU time and team ownership.
- ▌ Mlsys Submission · brycewang-stanfordUse when auditing an MLSys submission for OpenReview readiness, covering the two-column 10-page body excluding references, the separate appendix upload, double-blind rules that still permit arXiv posting, research-versus-industrial track requirements, dual-submission exceptions, style-kit compliance, and desk-reject triage before the deadline.
- ▌ Mlsys Experiments · brycewang-stanfordUse when designing or auditing the evaluation of an MLSys paper, selecting representative workloads and hardware, tuning baselines symmetrically, reporting throughput, latency tails, memory, cost, and quality together, structuring ablations that attribute gains to mechanisms, and building scaling and sensitivity evidence reviewers trust.
- ▌ Mlsys Camera Ready · brycewang-stanfordUse when preparing an accepted MLSys paper for publication on proceedings.mlsys.org, covering de-anonymization of the research-track PDF, restoring company and repository identity, reconciling promised rebuttal edits, the reserved-ticket registration window for authors, and sequencing camera-ready work against the artifact-evaluation deadline.
- ▌ Mlsys Related Work · brycewang-stanfordUse when positioning an MLSys submission against the fast-moving ML-systems literature scattered across OSDI, SOSP, NSDI, ASPLOS, and ML venues, handling arXiv-first and open-source-first prior work, comparing against production systems that have no paper, and writing the delta statement two reviewer cultures will both accept.
- ▌ Mlsys Supplementary · brycewang-stanfordUse when deciding what goes into an MLSys appendix versus the 10-page body, exploiting the venue's unlimited separately-uploaded appendix that reviewers are not required to read, organizing configs, traces, extended results, and anonymized code pointers, and keeping supplementary material blinded for the research track.
- ▌ Mlsys Writing Style · brycewang-stanfordUse when revising an MLSys paper's prose and structure, building the measured-bottleneck opening, naming the mechanism instead of listing optimizations, writing evaluation sections as answers to research questions, quantifying every performance claim with workload context, and fitting the argument into the venue's 10-page two-column body.
- ▌ Mlsys Review Process · brycewang-stanfordUse when reasoning about how MLSys peer review works, covering the OpenReview workflow, the mixed ML-and-systems reviewer pool and how each half scores differently, the compressed response window, industrial-track review expectations, decision dynamics, and what the post-acceptance artifact stage means for review strategy.
- ▌ Mlsys Author Response · brycewang-stanfordUse when drafting MLSys author responses on OpenReview under the venue's compressed rebuttal window, prioritizing systems-reviewer objections about workload representativeness, baseline tuning, and missing measurements, deciding whether to run new experiments in days, and keeping replies anonymous and decision-focused.
- ▌ Mlsys Reproducibility · brycewang-stanfordUse when hardening the reproducibility of MLSys performance claims, pinning the full system layer from driver to interconnect, separating ML randomness from systems noise, choosing repetition counts and variance reporting for throughput and latency numbers, and disclosing hardware, workloads, and cost so strangers can re-measure results.
- ▌ Mlsys Topic Selection · brycewang-stanfordUse when deciding whether a project belongs at MLSys rather than OSDI, SOSP, NSDI, ASPLOS, ATC, EuroSys, or the ML conferences, testing for genuine ML-systems co-design, choosing between the research and industrial tracks, and sharpening the systems-for-ML or ML-for-systems framing before writing starts.
- ▌ Mlsys Artifact Evaluation · brycewang-stanfordUse when packaging an accepted MLSys paper's code, configs, and measurement scripts for the venue's post-acceptance artifact evaluation, targeting the Availability, Functional, and Reproducible badges, writing the Artifact Appendix, handling hardware that AE reviewers cannot access, and answering anonymous evaluator questions.