spb/wp13_uqo

1\begin{table}[htbp]2\centering3\caption{Operational characteristics of the platform. Latencies are indicative order-of-magnitude4figures for the corresponding subsystems rather than benchmarked measurements; they motivate the5streaming and single-call design choices discussed in the text.}6\label{tab:robustness}7\begin{threeparttable}8\begin{tabular}{@{}lll@{}}9\toprule10Dimension & Mechanism & Indicative characteristic \\11\midrule12First-token latency   & SSE streaming                 & sub-second perceived response \\13Tool concurrency      & batched execution             & $\leq 5$ tools / batch, 500\,ms spacing \\14Context window        & adaptive summarization        & effectively unbounded history \\15Generation cap        & per-response limit            & 64{,}000 tokens \\16Python analysis       & sandboxed subprocess          & seconds per execution \\17PDF rendering         & headless browser              & several seconds \\18Slide compilation     & two-pass \LaTeX{}             & roughly ten seconds \\19Cost control          & local token estimation        & per-session accounting \\20Rate-limit safety     & inter-batch delay             & bounded provider load \\21\bottomrule22\end{tabular}23\begin{tablenotes}[flushleft]24\footnotesize25\item \textit{Notes.} Figures characterize the subsystems qualitatively to explain design26trade-offs; a controlled performance evaluation is left to future work27(Section~\ref{sec:conclusion}).28\end{tablenotes}29\end{threeparttable}30\end{table}31