Local-first chat-log forensics workbench: K-S, M/M/1, Barabási heavy-tail, Granger. Pure Python stdlib. Cited research.
-
Updated
May 7, 2026 - Python
Local-first chat-log forensics workbench: K-S, M/M/1, Barabási heavy-tail, Granger. Pure Python stdlib. Cited research.
Comparing LLM serving under real ShareGPT traces vs synthetic Poisson workloads. Real distributions produce 42x worse P99 TTFT and earlier SLO violations — scheduling policies only differentiate under heavy-tailed real traffic.
Add a description, image, and links to the heavy-tail topic page so that developers can more easily learn about it.
To associate your repository with the heavy-tail topic, visit your repo's landing page and select "manage topics."