research-paper-writing

Follow a phased pipeline: define a one-sentence contribution, map every experiment to a claim, run baselines with statistical significance tests, then draft using the venue's official LaTeX template. Complete the mandatory conference checklist before submission to avoid desk rejection.

Never generate BibTeX from memory; fetch it programmatically via DOI content negotiation from CrossRef or the Semantic Scholar API. Verify each paper exists in at least two sources and mark anything unverifiable as a placeholder citation.

Report error bars with the method specified, 95% confidence intervals via bootstrapping, and McNemar's test for pairwise method comparisons. Include effect sizes such as Cohen's h and state the number of runs or seeds.

Yes, papers missing the NeurIPS paper checklist face automatic desk rejection. The checklist has 16 items covering limitations, reproducibility, statistical significance, compute resources, and broader impacts, and appears after references outside the page limit.

Use human evaluation when automated metrics do not capture your claims, such as fluency, helpfulness, or preference in generation tasks at NLP venues. Report annotator counts, inter-annotator agreement like Krippendorff's alpha, and compensation details.

NeurIPS allows 9 pages, ICML 8 pages, ICLR 9 pages, ACL 8 pages for long papers, AAAI 7 pages, and COLM 9 pages for main content. References and appendices are unlimited at all of these venues.