Search papers, labs, and topics across Lattice.
This paper introduces Search-to-Record, a novel database-assurance task that leverages search-enabled large language models (LLMs) to reconstruct corporate-event records from public sources, addressing the critical need for financial institutions to detect inaccuracies in vendor databases. The authors present DelistBench, a benchmark consisting of 1,200 records focused on security-level delisting announcements, and evaluate five models under both closed-book and web-enabled conditions. Results show that web access significantly enhances accuracy, with the best-performing model achieving an overall joint accuracy of 81.5% within seven days, while low-cost systems can achieve competitive accuracy at a fraction of the API cost.
Web-enabled LLMs can boost corporate-event record accuracy by up to 48 percentage points, making them indispensable for financial institutions seeking reliable data.
Financial institutions need an independent way to detect missing, stale, and misclassified corporate-event records in vendor databases. We introduce Search-to-Record, a database-assurance task in which search-enabled large language models reconstruct institution-defined event records from public sources for a known security universe and historical cutoff, and DelistBench, a 1,200-record benchmark for security-level delisting announcements. We evaluate five models in paired closed-book and web-enabled conditions. Web access raises announcement-date accuracy within seven days by 34.0 to 48.0 percentage points and event-status accuracy by approximately 2.8 to 21.7 points; the best system achieves 81.5% overall joint accuracy within seven days. Economy web systems achieve 75.9-78.3% overall joint accuracy within seven days at 4.5-6.6% of the API cost of the most expensive web system. Risk-based triage identifies low-error subsets, although the highest-coverage operating point still sends 27.3% of the balanced test set to review. The evaluation identifies web retrieval as the main source of timing gains and shows that low-cost systems can approach the best system's accuracy. Together, Search-to-Record, DelistBench, and the evaluation provide concrete deployment guidance: calibrate triage to local event prevalence and market mix, preserve positive-event recall, and route positive and ambiguous cases to targeted review.
