mirror of
https://github.com/zoo-apps/papers.git
synced 2026-10-30 14:56:54 +00:00
zeekayandHanzo Dev
dced1dcfb9
Bylines, paper authorship, correspondence addresses and the founder blurb under each post named an individual. They name the organisation that published the work instead: the brand on posts and pages, the research byline on papers, and research@ in place of a personal address. Both are what these repositories already used elsewhere — nothing new was coined. Co-authors are untouched. Where a paper named someone else alongside, they stay named; only the one byline changed. Bibliography entries, running heads, PDF metadata and affiliations move with it, so a citation still resolves to whoever the cited paper names. Sample and fixture data uses the placeholder name these tests already use elsewhere, so no assertion changed meaning. Co-authored-by: Hanzo Dev <[email protected]>
77 lines
3.0 KiB
TeX
77 lines
3.0 KiB
TeX
\documentclass[11pt,letterpaper]{article}
|
|
\usepackage{shared/zoocover}
|
|
|
|
% Packages
|
|
\usepackage[utf8]{inputenc}
|
|
\usepackage[T1]{fontenc}
|
|
\usepackage{amsmath,amssymb,amsthm}
|
|
\usepackage{algorithm}
|
|
\usepackage{algorithmic}
|
|
\usepackage{graphicx}
|
|
\usepackage{hyperref}
|
|
\usepackage{xcolor}
|
|
\usepackage{booktabs}
|
|
\usepackage{tikz}
|
|
\usepackage{pgfplots}
|
|
\pgfplotsset{compat=1.18}
|
|
\usetikzlibrary{shapes,arrows,positioning,calc}
|
|
\usepackage{listings}
|
|
\usepackage{caption}
|
|
\usepackage{subcaption}
|
|
% geometry loaded by zoocover
|
|
|
|
% Theorem environments
|
|
\newtheorem{theorem}{Theorem}[section]
|
|
\newtheorem{lemma}[theorem]{Lemma}
|
|
\newtheorem{proposition}[theorem]{Proposition}
|
|
\newtheorem{corollary}[theorem]{Corollary}
|
|
\newtheorem{definition}{Definition}[section]
|
|
|
|
% Custom commands
|
|
\newcommand{\Zoo}{\textsc{Zoo}}
|
|
\newcommand{\Lux}{\textsc{Lux}}
|
|
\newcommand{\PoAI}{\textsc{PoAI}}
|
|
\newcommand{\Quasar}{\textsc{Quasar}}
|
|
\newcommand{\TEE}{\textsc{tee}}
|
|
|
|
% Hyperref setup
|
|
\hypersetup{
|
|
colorlinks=true,
|
|
linkcolor=blue,
|
|
citecolor=blue,
|
|
urlcolor=blue
|
|
}
|
|
|
|
\title{\textbf{Zoo Network Consensus: Quasar + PoAI Dual-Layer Verification}}
|
|
\author{
|
|
Zoo Labs Foundation Research\thanks{[email protected]}\\
|
|
\textit{Zoo Labs Foundation}\\
|
|
\texttt{[email protected]}
|
|
}
|
|
|
|
\date{November 2025}
|
|
|
|
\begin{document}
|
|
|
|
\zoocoverpage
|
|
|
|
\begin{abstract}
|
|
We present the dual-layer consensus architecture of the Zoo Network (chain ID 200200): \Quasar{} for block consensus and \PoAI{} for AI compute verification. \Quasar{}, inherited from the Lux Network, provides deterministic finality in 10ms with $O(n)$ message complexity, making it the fastest BFT consensus protocol deployed in production. \PoAI{} operates as a verification overlay that attests to the correctness of AI compute contributions---inference, training, and data processing---performed by Zoo Network validators. The dual-layer design separates concerns: \Quasar{} handles transaction ordering and block production with sub-second finality, while \PoAI{} handles compute quality assessment with epoch-based evaluation. We benchmark the combined system on Zoo testnet with 21 validators and 500 GPU workers, demonstrating 10.9ms combined finality (10ms \Quasar{} + 0.9ms \PoAI{} overhead), 4.76M TPS sustained throughput, and 98.2\% of validator compute devoted to useful AI work. We compare against single-layer alternatives and show that the dual-layer approach achieves 15\% higher useful compute utilization without sacrificing consensus safety or liveness.
|
|
|
|
\textbf{Keywords}: consensus protocol, Quasar, Proof of AI, dual-layer verification, Zoo Network
|
|
\end{abstract}
|
|
|
|
\input{sections/01-introduction}
|
|
\input{sections/02-background}
|
|
\input{sections/03-dual-layer-design}
|
|
\input{sections/04-quasar-benchmarks}
|
|
\input{sections/05-poai-overhead-analysis}
|
|
\input{sections/06-tee-attestation-for-gpu-providers}
|
|
\input{sections/07-comparative-evaluation}
|
|
\input{sections/08-discussion}
|
|
\input{sections/09-related-work}
|
|
\input{sections/10-conclusion}
|
|
\input{sections/99-acknowledgments}
|
|
|
|
\end{document}
|