spb/firc0001_uqo

1% Generated by code/make_tables.py from results/grounding/grounding\_by\_group.csv, grounding\_element\_level.csv, grounding\_all\_structured.csv, fallback\_bounds.json. Do not edit by hand.2\begin{table}[H]3\centering\footnotesize4\setlength{\tabcolsep}{4pt}5\caption{Mechanical grounding of the structured fields of all stored mentions in the stored section text}\label{tab:grounding}6\begin{threeparttable}7\begin{adjustbox}{max width=\textwidth}8\begin{tabular}{lrrrrr}9\toprule10\multicolumn{6}{l}{\itshape Panel A: share of mentions whose field is found literally in the stored section text (\%, 95\% CI)} \\11Field & Mentions with the field & All forms & 10-K & 10-Q & 8-K \\12\midrule13Cited years & 252{,}469 & 99.8 [99.7, 99.8] & 99.6 & 99.8 & 99.8 \\14Partners & 81{,}320 & 99.4 [99.3, 99.5] & 99.3 & 99.1 & 99.6 \\15Amount (number) & 120{,}240 & 93.7 [93.6, 93.9] & 97.4 & 97.5 & 86.9 \\16Amount (number + currency token within 40 chars) & 120{,}240 & 80.5 [80.2, 80.7] & 86.8 & 82.3 & 74.9 \\17Locations & 118{,}783 & 78.3 [78.0, 78.6] & 85.7 & 82.1 & 67.8 \\18Technology tags & 36{,}740 & 63.6 [63.0, 64.2] & 63.8 & 64.8 & 58.2 \\19\midrule20\multicolumn{6}{l}{\itshape Panel B: element level (every listed year, partner, location or tag counted separately)} \\21Partners & 122{,}590 elements & 99.6 & & & \\22Locations & 181{,}898 elements & 84.2 & & & \\23Technology tags & 47{,}361 elements & 70.6 & & & \\24Cited years & 307{,}399 elements & 99.7 & & & \\25\midrule26\multicolumn{6}{l}{\itshape Panel C: mentions with every structured field grounded (\%), by form, run and confidence bin} \\27All mentions with at least one structured field & 276{,}168 & 88.2 & 92.2 & 92.4 & 76.6 \\28By run: June (backfilled) / October & 36{,}806 / 239{,}362 & 84.5 / 88.7 & & & \\29\quad confidence bin 0.45-0.6 & 696 & 90.7 & & & \\30\quad confidence bin 0.6-0.7 & 1{,}726 & 93.5 & & & \\31\quad confidence bin 0.7-0.8 & 13{,}313 & 94.2 & & & \\32\quad confidence bin 0.8-0.9 & 58{,}586 & 92.7 & & & \\33\quad confidence bin 0.9-1.0 & 201{,}847 & 86.4 & & & \\34\quad confidence bin 10-K & 69{,}065 & 92.2 & & & \\35\quad confidence bin 10-Q & 134{,}391 & 92.4 & & & \\36\quad confidence bin 8-K & 72{,}712 & 76.6 & & & \\37\bottomrule38\end{tabular}39\end{adjustbox}40\begin{tablenotes}\footnotesize41\item Population: all 308{,}065 stored mentions, joined to the text of their section as stored (legacy June text for the S\&P~500 10-Ks). A field is ``grounded'' when every element of the list (years, partners, locations, technologies) or the amount (tested in units, thousands, millions and billions with one to three significant digits) occurs literally in that text; a currency token is \$, USD, dollar, million or billion within 40 characters of the number. Before the search, element and text are normalised identically: case-folded, punctuation replaced by spaces, whitespace collapsed, and a short alias list applied to the element (U.S., USA $\to$ united states; U.K. $\to$ united kingdom; UAE; EU); the element must then occur as whole words. The normalisation matters for partners: on a reservoir sample of 20{,}000 mentions (5{,}287 with a partner), the rate is 99.3\% under this rule and 96.1\% under a plain case-folded substring search (\texttt{results/grounding/partners\_normalisation\_check.json}). Confidence intervals are percentile bootstraps with 1{,}000 replicates clustered by section. Grounding is a necessary condition for correctness, not a measure of it: a grounded year can belong to another initiative, and an ungrounded location can be a correct normalisation (``Texas'' for ``Austin''); precision and recall require the human labels of gate (d). Panel C counts mentions with at least one structured field. Upper bounds on the regex back-fill from the model's own paraphrase (\texttt{llm.py}): the stored years equal the regex output on the description for 98{,}438 of 252{,}469 mentions with years (39\%) and the technology list for 16{,}949 of 36{,}740 (46\%); 604 mentions cite a year absent from the section text.42\end{tablenotes}43\end{threeparttable}44\end{table}45