How submission works¶
What you do, what MLCommons does, what reviewers do, and when your results become public.
Overview¶
flowchart TD
A[Plan the curve<br/>choose C_max, compute regions] --> B[Run 7–32 measurement points<br/>Offline included]
B --> C[Author system_desc.json and point.yaml<br/>per point, system_power.json per system]
C --> D[Register each run<br/>runs create]
D --> E[Assemble + validate<br/>submissions create]
E -->|checker errors| C
E -->|checker passes| F[Bundle uploaded<br/>peer-review PR opened]
F --> G[Week 0<br/>automated compliance]
G -->|fails| X([Rejected<br/>correct and resubmit as new]):::bad
G -->|passes| H[Weeks 1-3<br/>peer review, objections filed]
H --> I[Weeks 4-6<br/>objection resolution]
H -->|all objections resolved| K
I -->|objection needs a fix| C
I -->|unresolved at Week 6| J[Dispute resolution<br/>~5 weeks, chair decides]
I --> K[Finalized<br/>peer review pending tag removed]
J --> K
H -.->|no response for<br/>10 business days| Y([Withdrawn]):::bad
K --> L[Published in the next cohort<br/>1st or 3rd Wednesday]
classDef bad stroke:#c62828,stroke-width:2px;
Rolling submission and cohorts¶
There is no submission deadline: submit on any day and the submission enters the pipeline on receipt
(§4.1). Results publish in cohorts on the 1st and 3rd Wednesday of each month
(YYYY-MM-C0 and YYYY-MM-C1, §4.2).
To make a cohort, your submission has to pass automated compliance at least one business day before that Wednesday, and every review clock counts from the cohort you first appear in, not the day you uploaded (§4.3).
The review phases¶
The windows, from §6.5:
| Phase | Window | What happens |
|---|---|---|
| Automated compliance | Week 0 (may finish in a day) | Any failure rejects the submission; you resubmit as a new one |
| Peer review | Weeks 1–3 | Objections filed as GitHub issues. None new after Week 3. Resolve everything here for early finalization |
| Objection resolution | Weeks 4–6 | Carried-over objections must be resolved |
| Dispute resolution | From Week 6, ~5 weeks | Automatic escalation for anything still open; the chair decides |
Your response deadline¶
This is easy to miss. Once someone files an objection against you, you have 3 business days to post an initial response, either a resolution schedule or a counter-argument with evidence (§6.3). Miss it and penalties apply automatically. They add up and can't be reversed:
| Business days without your response | Penalty |
|---|---|
| 3 | Finalization delayed by 1 cohort |
| 6 | Delayed by 2 cohorts |
| 10 | Submission is withdrawn |
See After you submit.
Who reviews you¶
Organisations with a finalized Endpoints result in the lookback window form the review committee (§2.1), and each submission also gets one randomly assigned reviewer from that pool, never from your own organisation (§2.6). Being a competitor is not a conflict of interest (§2.4).
When your results become public¶
You pick one of three publication modes when you submit, and you can't change it afterwards (§6.2):
| Mode | Public before finalization? | Peer review starts | Finalized result goes public |
|---|---|---|---|
| Confidential review (default) | No | When automated checks pass | The first cohort after finalization |
| Confidential review, embargoed | No | When automated checks pass | On your embargo date |
| Provisional publication | Yes, tagged "peer review pending" | At provisional publication | The tag is removed at finalization |
Use the embargoed mode to time a launch without early numbers going out, and provisional publication when you need numbers public before review ends. An embargo under provisional publication also delays the start of your review. The details of each mode are in §6.2.1–6.2.3, and who can see what during review is in §6.10.
After publication¶
Published results can't be edited; an error gets the affected points or the submission invalidated (§8.1). Others can challenge a result until the later of the next audit vote or 90 days after finalization, and you keep the system available for audit until then (Scope and Standing for Late Concerns).
Audits¶
MLCommons runs 8 audits a year: each quarter one drawn at random and one chosen by committee vote from nominations (§10). The thing to plan for: you have to keep the system as submitted from the moment it's nominated, not only once it's picked (§10.2). What an audit asks of you, and what happens if you fail, is in §10.5. What the auditor checks is in the MLPerf Endpoints Audit Guidelines.
Next: Divisions and scenarios
The policy repository wins
This page summarises policy in submitter language. It does not restate it. Where this
page and mlcommons/endpoints_policies
disagree, the policy repository is correct and this page is a bug. Please
report it.
Last verified against: mlcommons/endpoints_policies@v1.0_rules_dev (6b0b1ef), 2026-09-24.