← all publishers

Alexyskoutnev

@alexyskoutnev source repo

10 published skills

  1. Method Writing · alexyskoutnev bundle
    Write or revise a paper's Method section using recipes distilled from dissecting top ML researchers' papers. Use when drafting or restructuring a method section, choosing its architecture figure, algorithm box, or notation, deciding how to introduce and defend a system's components, or fixing a method that reads as a flat wall of mechanics. Applies the N-named-parts skeleton, the setup-banks-notation rule, the star-component discipline, the running-example rule, and mechanical checklists.
    0
    installs
  2. Sentence Audit · alexyskoutnev
    Sentence-by-sentence precision pass for paper prose, divide a paragraph into sentences, test each against the first-time-reader defect list, rewrite minimally, condense aggressively; plus the LaTeX editing invariants (wrapping, per-paragraph commits, layout checks) that keep an edit pass safe. Use when polishing a section or caption for submission, when a paragraph reads wrong but the fault is unclear, or when auditing a figure-caption-table exhibit as a unit.
    0
    installs
  3. Caption Writing · alexyskoutnev bundle
    Write or revise figure, table, and algorithm captions using recipes distilled from dissecting top ML researchers' papers. Use when adding or editing any caption, when a new exhibit enters a paper, or when a caption fails to stand on its own. Applies the name-decode-claim structure, the flagship-claims/workhorse-describes split, self-containment over deduplication, paired-number discipline, and mechanical checklists.
    0
    installs
  4. Figure Creation · alexyskoutnev bundle
    Design and build publication figures for ML venues using the design decisions and the mechanical export steps that accepted NeurIPS, ICML, and ICLR papers follow. Use when a paper needs a new figure, when a system diagram is authored in Figma, when a data plot is built in matplotlib, or when an existing figure must be audited for column width, font size, vector output, and color accessibility before submission. Applies the one-job rule, the form heuristic, the two authoring tracks, the never-smaller-than-the-caption rule, redundant encoding, and the pre-submission checklist.
    0
    installs
  5. Results Writing · alexyskoutnev bundle
    Write or revise a paper's Experimental Results section using recipes distilled from dissecting top ML researchers' papers. Use when drafting or editing a results section, its tables, figures, or captions, an ablation, or a limitations subsection, or when a results section reads as an undifferentiated scoreboard. Applies the claim-headed paragraph rule, the measurement-over-scoreboard weighting, the named-component lesion, paired-number discipline, and mechanical checklists.
    0
    installs
  6. Abstract Writing · alexyskoutnev bundle
    Write or revise a research-paper abstract using three master recipes distilled from dissecting top ML researchers' abstracts. Use when drafting an abstract, revising one for submission, or when the user asks how to structure or tighten an abstract. Applies the six-questions structure, the three recipes, and a mechanical checklist (short sentences, contribution first, no "can", no bare "This", numbers in pairs).
    0
    installs
  7. Rebuttal Writing · alexyskoutnev
    Plan and write an academic rebuttal in two stages. Stage 1 triages the reviews, split them into atomic concerns, diagnose the doubt behind each, rate severity and decision impact, and produce a prioritized P0-P3 experiment plan. Stage 2 drafts the rebuttal from completed results using Direct Answer -> Evidence -> Revision. Use when reviews arrive, when deciding whether a rebuttal is worth the author's time, when ranking rebuttal experiments, or when drafting or quality-checking the response. Never fabricates results, reviewer intent, or manuscript locations.
    0
    installs
  8. Conclusion Writing · alexyskoutnev bundle
    Write or revise a paper's Conclusion using recipes distilled from dissecting top ML researchers' papers. Use when drafting or editing a conclusion, or when applying feedback to it. Applies the three-paragraph envoi shape, the reprise-not-report rule, the flip-the-shortfall move, the refutable-vista close, and mechanical checklists.
    0
    installs
  9. Supplement Writing · alexyskoutnev bundle
    Write or revise a paper's supplement/appendix using a method distilled from dissecting seven appendices by leading ML researchers. Use when drafting or restructuring a supplement, adding an exhibit (prompt gallery, learned-artifact listing, qualitative trace, failure section, protocol tables), or when deciding what belongs in the appendix versus the main paper. Applies the evidence-locker rule, the prose-first exhibits-last order, verbatim-artifact discipline, and mechanical checklists.
    0
    installs
  10. Intro Related Writing · alexyskoutnev bundle
    Write or revise a paper's Introduction and Related Work sections using recipes distilled from dissecting top ML researchers' papers. Use when drafting an intro, a contributions list, or a related-work section, or when restructuring how a paper positions against rivals. Applies the seven-question funnel, the credit-then-contrast pattern, and mechanical checklists.
    0
    installs