PianoFingerprints: Rediscovering Chopin via Composition-Aware Contrastive Learning

Music discovery from symbolic archives requires relationships among works that persist across heterogeneous collections and MIDI realizations. We propose PianoFingerprints, a composition-anchored multilevel atlas combining co-registered three-channel time-pitch MIDI tensors, ViT-Base, and source-priority multi-positive contrastive learning. A segment-realization-work hierarchy separates reproducible work structure from archive and realization variation. Across 47,489 realizations from 11 archives, KDF is held out from training for cross-archive evaluation and fitted separately for single-archive discovery. Held-out KDF contains 232 nonempty realizations of 139 normalized works. The performance-archive geometry agrees with work geometry independently aggregated from held-out KDF ($r_s=0.767$), showing that relative work similarities recur across archives. Two nonoverlapping halves of the KDF realizations also yield similar work geometries ($r_s=0.870$), indicating stability across different realizations of the same works. In the single-archive atlas, mean within-genre similarity exceeds mean cross-genre similarity ($0.199$ versus $0.045$). There is no evidence that work distinctiveness has a statistically significant global linear association with composition year ($r=-0.148$, $p=0.075$). For 66 cross-genre neighbour pairs reproduced in both archive views, the closest ten-second passages have a mean symbolic similarity of $0.548$. Shuffling specific work identities within each genre, while retaining the same pair of genres for every relation, yields a permutation mean of $0.403$ ($\Delta =0.145$, $p_{\mathrm{perm}}=1.0\times 10^{-4}$). Together, these results position PianoFingerprints as a corpus-conditioned discovery instrument for Chopin archives, using cross-archive replication to identify reproducible structure and local passages to support interpretation.