research-paper-writing

Automate ML research paper writing by orchestrating literature review, experiment design, drafting, revision, and conference submission. This pipeline manages iterative cycles from idea generation to final manuscript, targeting ML/AI conferences like NeurIPS, ICML, and ICLR.

Yes, you can verify BibTeX citations programmatically to prevent hallucinated references. The workflow retrieves and validates citation entries automatically using libraries like semanticscholar, arxiv, and habanero to ensure your references are accurate.

Generate reproducibility artifacts by automatically attaching logs, data splits, and configuration files to your project. This workflow enforces reproducibility by capturing these artifacts during experiment design and analysis using dependencies like scipy, numpy, and matplotlib.

Structure your pipeline by implementing a structured metadata schema and frontmatter requirements for your project. This end-to-end workflow integrates drafting and revision with conference submission formatting for NeurIPS, ACL, ICML, ICLR, AAAI, and COLM.

Yes, you need Python scientific libraries like scipy, numpy, and matplotlib to support the pipeline. These dependencies are required for experiment design, data analysis, and generating reproducibility artifacts with SciencePlots visualizations.

Yes, the paper-writing workflow supports iterative drafting cycles before final conference submission. You initiate a project, set milestones, and iterate on drafts, allowing continuous revision from initial idea through final manuscript and submission.