FINDING #3720 · UNIT ID 818558903
pyc-team/pytorch_concepts
PyC (Pytorch Concepts) is a PyTorch-based library for designing concept-based interpretable deep learning models.
SURPRISE SCORE
0.00
Score Breakdown
SCORE = SURPRISE × ENGAGEMENT^0.7 × FRESHNESS × VISIBILITY × CONFIDENCE
SURPRISE = WINDOW STARS / DAYS / (AUDIENCE + FLOOR)
68% OF STARS IN ARCHIVE
[BOT] SUSPECTED STAR BOT — SCORE PENALIZED. SIGNATURES:
S1a · OWNER ACCOUNT AND REPO CREATED SAME DAY
S2 · NO EXTERNAL ISSUE/PR AUTHORS DESPITE 100+ STARS
S5 · PREDATES WINDOW, YET HALF+ OF ALL ITS STARS LANDED IN IT
S7 · THIS REPO IS ~ALL OF THE OWNER'S STARS
Growth Telemetry
VELOCITY /D
34.33
ACCEL
-38.50
RETENTION
14.4%
PEAK 2026-08-27 · FORK-RETENTION 42.9% · 103 STARS / WINDOW
Author Audience
AUDIENCE
41
FOLLOWERS
5
OWNER ★
157
Engagement Signals
FORKS
21
ISSUE AUTH
0
PR AUTH
0
UNIQUE STARGAZERS 103 / 103 (DIVERSITY 1.00)
Why This Is A Finding
pyc-team/pytorch_concepts собрал 103 звёзд за окно, тогда как у автора всего 5 подписчиков — эффективная аудитория ≈ 41. Это даёт surprise-индекс 0.42 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 42.9% и 0 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.
Related Findings
RANKS ABOVE 59% OF 9102 FINDINGS
METRICS IN CONTEXT
MEDIAN ACROSS ALL 9102 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW
METRICVALUEMEDIANΔ MEDPERCENTILE
SCORE0.000.00+0.00ABOVE 59%
VELOCITY34.335.67+28.67ABOVE 92%
RETENTION14.4%45.5%-31.1 PPABOVE 26%
FORKS21340-318ABOVE 12%
SURPRISE0.420.00+0.42ABOVE 99%