|
| 1 | +# Downstream review concerns (address in this revision) |
| 2 | + |
| 3 | +## MAINTAINER DECISION (apply verbatim) |
| 4 | + |
| 5 | +The spec panel flagged a methodology drift between this idea and the spec: |
| 6 | +the idea promises "predictive power", an 80/20 train/test split, |
| 7 | +cross-validated R², ANOVA, and hold-out validation, while the spec treats |
| 8 | +the dataset as a complete census analyzed descriptively. This decision |
| 9 | +settles the question permanently. Reviewers and revisers MUST NOT re-raise |
| 10 | +it. |
| 11 | + |
| 12 | +1. **Descriptive framing is FINAL.** The dataset is the complete census of |
| 13 | + hyperbolic prime knots with ≤ 13 crossings — there is no larger |
| 14 | + population to generalize to, so hold-out validation is not applicable. |
| 15 | + Remove from this idea (everywhere, including the literature-gap |
| 16 | + statements and methodology sketch): the 80/20 train/test split, |
| 17 | + cross-validated R² model selection, ANOVA, "statistical power |
| 18 | + assessment", and "hold-out validation". Replace with: |
| 19 | + - Model assessment via goodness-of-fit on the full census: R², AIC/BIC, |
| 20 | + and MAE. |
| 21 | + - Group comparison (alternating vs. non-alternating) via descriptive |
| 22 | + statistics: mean differences and variance ratios (exact population |
| 23 | + parameters; no inferential testing). |
| 24 | + - Residual analysis identifying hyperbolic knot families deviating by |
| 25 | + ≥ 2 standard deviations from the fitted trend. |
| 26 | +2. **Restate the research question** as: "To what extent do crossing number |
| 27 | + and braid index jointly explain variation in hyperbolic volume across |
| 28 | + the complete census of hyperbolic prime knots with ≤ 13 crossings, |
| 29 | + stratified by alternating/non-alternating classification?" Use |
| 30 | + "explain/account for variance" language; do not frame claims as |
| 31 | + predictive performance requiring held-out data. |
| 32 | +3. **Phase 1 invariant boundary is unchanged**: crossing number, braid |
| 33 | + index, and hyperbolic volume ONLY. Additional invariants (arc index, |
| 34 | + Seifert circle count, bridge number) remain excluded from Phase 1 and |
| 35 | + reserved for Phase 2+ exploratory extension after primary results are |
| 36 | + established. The spec deferring them to Phase 2+ is CONSISTENT with |
| 37 | + this boundary — reviewers must not flag it as scope creep. |
| 38 | +4. **Citations must be verbatim-verified.** Keep every existing verified |
| 39 | + citation exactly as written. Never introduce placeholder citation |
| 40 | + fragments (e.g. "[arXiv identifier]"); if a specific identifier is not |
| 41 | + in the verified list, omit the citation rather than approximating it. |
| 42 | + State each distinct source citation once per section — do not repeat |
| 43 | + the same URL multiple times in a single paragraph. |
| 44 | + |
| 45 | +Everything else in the idea (hyperbolic-prime-knot restriction, dataset |
| 46 | +sources, validation of computed invariants against KnotInfo, GitHub |
| 47 | +Actions compute envelope of ≤ 6 h / ≤ 7 GB RAM / 2 CPU cores) is correct |
| 48 | +and must be preserved verbatim. |
0 commit comments