chrisliu298/awesome-on-policy-distillation — GitHub trend

FINDING #4754 · UNIT ID 1189057629

chrisliu298/awesome-on-policy-distillation

A curated collection of papers, technical reports, frameworks, and tools for on-policy distillation (OPD) of large language models

[ ][ GITHUB ↗ ]

SURPRISE SCORE

0.00

Score Breakdown

SURPRISE0.00539

ENGAGEMENT0.33

FRESHNESS1.13

SCORE = SURPRISE × ENGAGEMENT^0.7 × FRESHNESS × VISIBILITY × CONFIDENCE

SURPRISE = WINDOW STARS / DAYS / (AUDIENCE + FLOOR)

5% OF STARS IN ARCHIVE

Growth Telemetry

VELOCITY /D

1.60

ACCEL

-0.05

RETENTION

29.5%

PEAK 2026-09-26 · FORK-RETENTION 0.0% · 48 STARS / WINDOW

Author Audience

AUDIENCE

257

FOLLOWERS

OWNER ★

1,806

Engagement Signals

FORKS

ISSUE AUTH

PR AUTH

UNIQUE STARGAZERS 48 / 48 (DIVERSITY 1.00)

Why This Is A Finding

chrisliu298/awesome-on-policy-distillation собрал 48 звёзд за окно, тогда как у автора всего 76 подписчиков — эффективная аудитория ≈ 257. Это даёт surprise-индекс 0.00539 (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 51% OF 9675 FINDINGS

#01Niko1221/Strata[ C++ ]0.73 #02Ebony-Vinyl/dsh-our-free-model[ JAVASCRIPT ]0.60 #03BlueHeisenberg/SCSKiller[ C# ]0.54 #04QingYunA/answer-me-with-html[ JAVASCRIPT ]0.51 #05Louis-CFM/coucou[ SWIFT ]0.50

METRICS IN CONTEXT

MEDIAN ACROSS ALL 9675 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW

METRICVALUEMEDIANΔ MEDPERCENTILE

SCORE0.000.00+0.00ABOVE 51%

VELOCITY1.603.80-2.20ABOVE 5%

RETENTION29.5%20.3%+9.3 PPABOVE 70%

FORKS43138-95ABOVE 26%

SURPRISE0.010.01-0.00ABOVE 43%