QwenLM/Qwen3-VL — GitHub trend

◂ BACK TO LEADERBOARD

FINDING #2357 · UNIT ID 849239437

QwenLM/Qwen3-VL

Qwen3-VL is the multimodal large language model series developed by Qwen team, Alibaba Cloud.

[ JUPYTER NOTEBOOK ][ ORG ][ GITHUB ↗ ]

SURPRISE SCORE

0.00

Score Breakdown

SURPRISE0.0000231

ENGAGEMENT0.87

FRESHNESS1.37

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

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

2% OF STARS IN ARCHIVE

Growth Telemetry

VELOCITY /D

1.72

ACCEL

-0.03

RETENTION

13.8%

PEAK 2026-02-28 · FORK-RETENTION 69.6% · 309 STARS / WINDOW

Author Audience

AUDIENCE

74,244

FOLLOWERS

18,219

OWNER ★

189,031

Engagement Signals

FORKS

1,833

ISSUE AUTH

PR AUTH

UNIQUE STARGAZERS 308 / 309 (DIVERSITY 1.00)

Why This Is A Finding

QwenLM/Qwen3-VL собрал 309 звёзд за окно, тогда как у автора всего 18,219 подписчиков — эффективная аудитория ≈ 74,244. Это даёт surprise-индекс 0.0000231 (звёзды относительно охвата автора, а не в абсолюте). Удержание форков 69.6% и 19 внешних контрибьюторов отделяют реальный инструмент от разовой вспышки. Акселерация отрицательная — внимание остывает после пика.

Related Findings

RANKS ABOVE 8% OF 2559 FINDINGS

#536duoan/TorchCode[ JUPYTER NOTEBOOK ]0.00335 #584CalvinXKY/InfraTech[ JUPYTER NOTEBOOK ]0.00302 #614GiovanniPasq/agentic-rag-for-dummies[ JUPYTER NOTEBOOK ]0.00285 #632Lordog/dive-into-llms[ JUPYTER NOTEBOOK ]0.00278 #653lyogavin/airllm[ JUPYTER NOTEBOOK ]0.0027

METRICS IN CONTEXT

MEDIAN ACROSS ALL 2559 FINDINGS · Δ vs MEDIAN · PERCENTILE = SHARE RANKED BELOW

METRICVALUEMEDIANΔ MEDPERCENTILE

SCORE0.000.00-0.00ABOVE 8%

VELOCITY1.722.88-1.17ABOVE 12%

RETENTION13.8%6.3%+7.5 PPABOVE 80%

FORKS1,8331,183+650ABOVE 61%

SURPRISE0.000.00-0.00ABOVE 7%