IJCAI Author Response
Use this after IJCAI reviews are released. Reopen the current Author's Response FAQ and
response template before drafting; rebuttal rules are cycle-specific.
Triage
- Respond only when a reviewer asks a pressing question, when a factual error could drive
rejection, or when an unethical review needs confidential escalation.
- Do not use the response to complain about tone, scores, reviewer identity, or broad
disagreement without a decision-relevant correction.
- Keep the public PDF response short. IJCAI-ECAI 2026 limited ordinary responses to one
page using the official response template.
- Do not upload new code, new links, new experiments, or new results. Clarifying figures or
examples are acceptable only when they explain material already in the submission.
- Use the confidential box only for unethical-review concerns; it is not visible to ordinary
PC reviewers.
- If a late review arrives during the response period, follow the current FAQ for extension
requests rather than guessing.
Compression pattern
- State the decision-critical correction.
- Anchor it to the submitted section, table, figure, theorem, appendix, or supplement.
- Answer the requested clarification and stop.
- If needed, add a camera-ready clarification promise without adding new evidence.
Triage table: respond or stay silent
The one-page budget forces ruthless selection. The response is not a dialogue, and phase-1
summary-reject papers usually get no response, so reserve the page for decision-changing items.
| Reviewer point |
Respond? |
How |
| Pressing question that could flip the score |
Yes |
Point to the submitted location |
| Factual error driving rejection |
Yes |
Correct precisely; anchor to section/table/theorem |
| "Add experiment X" |
No new results |
Promise a camera-ready note or cite existing evidence |
| Tone, score, broad disagreement |
No |
Do not spend budget relitigating |
| Suspected unethical review |
Confidential box |
Not the public PDF |
Worked vignette: a planning paper rebuttal
A planning paper draws one strong review and one that wrongly claims the optimality proof
assumes unit costs. That misreading could drive rejection, so it earns space: cite the
assumption already stated in the submitted theorem and point to the appendix proof, without a
new result. A second reviewer asks for an extra benchmark family; since no new experiments may
be added, promise a camera-ready note and reference the closest existing table. The complaint
about presentation tone gets no words.
Reviewer pushback and the venue-specific fix
- "You added new results in the response." Remove them; clarify only what the submission already
supports, since new evidence is disallowed.
- "Response argues the score, not a fact." Refocus on the one decision-critical correction.
- "Reviewer was unethical." Use the confidential channel, not the public response.
Output format
[Response budget] <pages used / current limit>
[Eligible reason] pressing question / factual error / unethical review
[Draft response] <IJCAI-ready text>
[Evidence anchor] <submitted location>
[Do-not-include list] <new results, code links, broad complaints, identity leaks>
1---2name: ijcai-author-response3description: Use when drafting IJCAI or IJCAI-ECAI author responses under the official one-page template, concise-response guidance, no-new-results constraints, code-link restrictions, factual-error triage, pressing-question triage, and confidential unethical-review channel.4---56# IJCAI Author Response78Use this after IJCAI reviews are released. Reopen the current Author's Response FAQ and9response template before drafting; rebuttal rules are cycle-specific.1011## Triage1213- Respond only when a reviewer asks a pressing question, when a factual error could drive14 rejection, or when an unethical review needs confidential escalation.15- Do not use the response to complain about tone, scores, reviewer identity, or broad16 disagreement without a decision-relevant correction.17- Keep the public PDF response short. IJCAI-ECAI 2026 limited ordinary responses to one18 page using the official response template.19- Do not upload new code, new links, new experiments, or new results. Clarifying figures or20 examples are acceptable only when they explain material already in the submission.21- Use the confidential box only for unethical-review concerns; it is not visible to ordinary22 PC reviewers.23- If a late review arrives during the response period, follow the current FAQ for extension24 requests rather than guessing.2526## Compression pattern27281. State the decision-critical correction.292. Anchor it to the submitted section, table, figure, theorem, appendix, or supplement.303. Answer the requested clarification and stop.314. If needed, add a camera-ready clarification promise without adding new evidence.3233## Triage table: respond or stay silent3435The one-page budget forces ruthless selection. The response is not a dialogue, and phase-136summary-reject papers usually get no response, so reserve the page for decision-changing items.3738| Reviewer point | Respond? | How |39| --- | --- | --- |40| Pressing question that could flip the score | Yes | Point to the submitted location |41| Factual error driving rejection | Yes | Correct precisely; anchor to section/table/theorem |42| "Add experiment X" | No new results | Promise a camera-ready note or cite existing evidence |43| Tone, score, broad disagreement | No | Do not spend budget relitigating |44| Suspected unethical review | Confidential box | Not the public PDF |4546## Worked vignette: a planning paper rebuttal4748A planning paper draws one strong review and one that wrongly claims the optimality proof49assumes unit costs. That misreading could drive rejection, so it earns space: cite the50assumption already stated in the submitted theorem and point to the appendix proof, without a51new result. A second reviewer asks for an extra benchmark family; since no new experiments may52be added, promise a camera-ready note and reference the closest existing table. The complaint53about presentation tone gets no words.5455## Reviewer pushback and the venue-specific fix5657- "You added new results in the response." Remove them; clarify only what the submission already58 supports, since new evidence is disallowed.59- "Response argues the score, not a fact." Refocus on the one decision-critical correction.60- "Reviewer was unethical." Use the confidential channel, not the public response.6162## Output format6364```text65[Response budget] <pages used / current limit>66[Eligible reason] pressing question / factual error / unethical review67[Draft response] <IJCAI-ready text>68[Evidence anchor] <submitted location>69[Do-not-include list] <new results, code links, broad complaints, identity leaks>70```
brycewang-stanford