FINDING #614 · UNIT ID 1384094559
Contrastive-LM/CLM
Contrastive Language Models (CLMs) are a new class of System One model trained with a contrastive learning objective that connects states and actions.
SURPRISE SCORE
0.00
Score Breakdown
SCORE = SURPRISE × ENGAGEMENT^0.7 × FRESHNESS × VISIBILITY × CONFIDENCE
SURPRISE = WINDOW STARS / DAYS / (AUDIENCE + FLOOR)
2% OF STARS IN ARCHIVE
[BOT] SUSPECTED STAR BOT — SCORE PENALIZED. SIGNATURES:
S2 · NO EXTERNAL ISSUE/PR AUTHORS DESPITE 100+ STARS
S7 · THIS REPO IS ~ALL OF THE OWNER'S STARS
S9 · CREATED JUST BEFORE WINDOW, STARS INVISIBLE IN ARCHIVE
Growth Telemetry
VELOCITY /D
2.00
ACCEL
-0.01
RETENTION
16.7%
PEAK 2026-09-24 · FORK-RETENTION 0.0% · 60 STARS / WINDOW
Author Audience
AUDIENCE
689
FOLLOWERS
44
OWNER ★
3,007
Engagement Signals
FORKS
262
ISSUE AUTH
0
PR AUTH
0
UNIQUE STARGAZERS 60 / 60 (DIVERSITY 1.00)
Why This Is A Finding
Contrastive-LM/CLM собрал 60 звёзд за окно, тогда как у автора всего 44 подписчиков — эффективная аудитория ≈ 689. Это даёт surprise-индекс 0.00274 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 0.0% и 0 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.
Related Findings
RANKS ABOVE 7% OF 663 FINDINGS
METRICS IN CONTEXT
MEDIAN ACROSS ALL 663 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW
METRICVALUEMEDIANΔ MEDPERCENTILE
SCORE0.000.00-0.00ABOVE 7%
VELOCITY2.002.77-0.77ABOVE 28%
RETENTION16.7%11.1%+5.5 PPABOVE 73%
FORKS2621,805-1,543ABOVE 14%
SURPRISE0.000.00+0.00ABOVE 78%