regenerate figures as 650-dpi LZW TIFFs, flattened RGB (publication-ready)

add DATA_AVAILABILITY.md: what's included, public dataset/backbone sources, reproduce-via-HF-Jobs, DOI note

rarity-route verified: spanning reproduces 0.46, sits between population & concentration, redundancy RULED OUT (incr rank 0.75m); candidate closed form rejected (over-predicts). Paper #2 §2 = two routes: low-rank existence proof (2a) + measured rarity mechanism (2b)

verify two reviewer-probe claims: (1) measured lesion spectra REFUTE 'low internal rank' (RankMe 339>307) -> correct attribution to RARITY across papers #1/#2/NEGATIVE_RESULT; (2) verified MedDINOv3/DINOv3=RoPE vs DINOv2=learned-absolute, paper #3 §3 stated precisely

G1c FINO-style adapter: metadata guidance offsets SSL-globalization directionally (+0.014) but no CI-significant recovery; F1b not de-risked at adapter scale (honest)

position vs FINO (Meta FAIR, arXiv:2606.05107): orthogonal frozen-probe vs metadata adaptation; de-risks F1b; complementary experiment noted

consolidation: RESEARCH_OVERVIEW; figures w/ multi-seed error bars + cross-objective + closed-form law; polished 3 drafts (selector numbers, refs, figure callouts)

v4: concentration-steering NEGATIVE (globalization is entangling, not removable); multi-seed rigor (S1 exact law, S2 tail-gap selector regret 0.006); papers updated

research programs v2 (S1-S5) + v3 (F1-F4) + cross-objective mechanism; paper drafts #2/#3; specs; figures

add model-card frontmatter

covtoken: label-free lesion-subspace token economy (reframed) + gated eval + paper draft

initial commit