Itcs Author Response

ITCS Author Response

The defining fact of this skill: ITCS has no rebuttal and no author-response phase. Following theory-conference norms, the program committee forms its reviews and reaches accept/reject without any input from you in between (see itcs-review-process). There is also no Major Revision and no revise-and-resubmit. So "author response" at ITCS is not a rebuttal-writing task — it is (1) doing the rebuttal's work inside the submission, before the deadline, and (2) handling the situation after the decision lands. Both are covered here.

Part 1 — The rebuttal you write in advance (in the paper)

At a rebuttal venue you would hold arguments in reserve to answer reviewers. At ITCS there is no reserve turn, so every such argument must be in the submitted PDF. Pre-empt the three objections an ITCS PC reliably raises:

• "Is the model trivial or all-powerful?" Answer with an anchoring result in the introduction — a separation or surprising possibility showing the model is neither empty nor everything (see itcs-writing-style).
• "Is this actually new?" Answer with a delta-first related-work paragraph naming the closest prior model and what it did not ask; search the ITCS/ICS back-catalog hardest, since that is where "already asked" hides (see itcs-related-work).
• "Why should the community care?" Answer with a concrete payoff — a connection, a reframing, an application sketch — on the first page.

Also disarm the objections you can foresee about your own paper:

• State scope and limitations yourself. A candidly owned gap is a strength at ITCS; the same gap discovered by a reviewer, with no rebuttal to contextualize it, is a weakness.
• Prove the load-bearing lemma in full, in the PDF's appendix — a reviewer who gets stuck on a deferred proof cannot ask you to clarify (see itcs-reproducibility).
• Add a "why not the obvious approach" remark where a reviewer might think your problem is easy; this is the single most common unspoken objection to a new-model paper.

Part 2 — After the decision

On accept

• Move to itcs-camera-ready: switch to lipics-v2021, de-anonymize, complete LIPIcs metadata, hit the production deadline.
• Reviews may still contain requests or corrections even without a rebuttal round. Fold reasonable fixes into the camera-ready — a corrected claim, a missing citation, a clarified definition — using the extra latitude the (unlimited) LIPIcs length allows.
• Register a presenter; consider a Graduating Bits talk if a coauthor is near graduation.

On reject

There is no resubmission to a later ITCS round this year (single annual cycle), so the move is to use the reviews, not to answer them:

• Mine the reviews for the real objection. Since they were written without expecting a reply, they tend to be candid. Identify whether the paper was judged not novel enough (a framing problem) or not correct/complete (a proof problem) — the fix differs sharply.
• Reroute by the diagnosis. If reviewers admired the depth but not the novelty, a depth venue (STOC/FOCS/SODA) may fit; if they doubted the novelty of the question, rethink the framing before resubmitting anywhere (see itcs-topic-selection).
• Fix proof gaps before the next submission, and update the arXiv/ECCC/ePrint full version so the durable record is correct.
• Do not email the PC to argue the decision. There is no appeal-by-rebuttal culture; a polite clarifying question to the chairs is the ceiling, and rarely changes an outcome.

What "author response" is NOT at ITCS

Output format

[ITCS response mode] pre-submission pre-emption / post-accept / post-reject
[Pre-emption] anchoring result? novelty delta? payoff? scope owned? load-bearing proof in PDF? (yes/no each)
[Post-accept] camera-ready fixes folded in? presenter registered? Graduating Bits considered?
[Post-reject] objection diagnosed (novelty vs correctness)? reroute target? full version corrected?
[Next action] <single clearest step>

Source: brycewang-stanford/Awesome-Journal-Skills → ITCS-Skills/skills/itcs-author-response/SKILL.md