Use this when splitting content between the body and the appendix. S&P's model is stricter than most: there is no separate supplement upload in the verified CFPs — everything is one PDF of at most 13 body pages plus at most 5 pages of references and appendices, 18 total, with all content past page 13 clearly marked as appendix and reviewers explicitly not required to read appendices (sp2027.ieee-security.org/cfpapers.html, checked 2026-07-08).
If the acceptance case needs it, it lives on pages 1–13. The appendix is for material a convinced reviewer might verify, never for material that does the convincing.
| Content | Placement | Why |
|---|---|---|
| Threat model, all of it | Body | The claim is undefined without it |
| Adaptive-attack evaluation summary | Body | The defense's acceptance case |
| End-to-end exploit walkthrough (one target) | Body | Existence proof of the attack |
| Per-target breakdown across 30 targets | Appendix | Verification depth, not persuasion |
| Full protocol/proof details | Appendix, with theorem statements in body | Statements are claims; derivations are checks |
| Ethics considerations | Body (submission); dedicated section added at camera-ready doesn't count against limits | Reviewed substance |
| Extended related work | Appendix | Positioning core stays in body (ieeesp-related-work) |
| Measurement instrument details (query lists, scan configs) | Appendix | Reproducibility depth |
References and appendices share the 5-page allowance. A security paper's bibliography routinely runs 2.5–3 pages in compsoc format, leaving perhaps two pages of real appendix. Consequences:
ieeesp-submission), including figure paths and dataset names
that identify a lab.Split audit — run over the section list:
for each body section: would a first-round reviewer cut the paper
without this? if no → appendix candidate
for each appendix item: does any *claim* in the body rely on it as
evidence? if yes → PROMOTE to body or weaken
the claim
page check: body ≤ 13 · refs+appendix ≤ 5 · total ≤ 18
· post-13 content labeled
A Revise expectations summary often asks for material authors originally appendixed ("bring the adaptive evaluation into the main text"). Cheapest prevention: at first submission, put the summary statistics of every appendix experiment in the body with a pointer down — reviewers who can see the number rarely demand the relocation.
[Page arithmetic] body <n>/13 · refs <n> + appendix <n> ≤ 5 · total <n>/18
[Marking] post-p13 labeled ✓/✗
[Promotion list] <appendix items the acceptance case secretly depends on>
[Demotion list] <body items a first-round reviewer would not miss>
[Anonymity] appendix swept ✓/✗