AI TrackerFrontier desk · AEST
Back to news
normalaitech

Testimony from Timelines: Evidence-Grounded Significance Testing of Temporal Clinical Hypotheses

arXiv cs.CL · 30 Oct 2026, 03:55 AEST

The study presents STRATA, a benchmark for evaluating significance testing in temporal clinical hypotheses, particularly in health contexts. It highlights the importance of decision-making in the testing process and introduces ART, a new paradigm for improving accuracy in model outputs.

Key facts

Read aloud

The study presents STRATA, a benchmark for evaluating significance testing in temporal clinical hypotheses, particularly in health contexts. It highlights the importance of decision-making in the testing process and introduces ART, a new paradigm for improving accuracy in model outputs. The verifier must specify the null, select the test and decide which records bear on the hypothesis, and any of these choices can go wrong while the conclusion still comes out right. We study the problem in health, where a wrong verdict is costly, and present STRATA, a benchmark that grades every decision on the way to a verdict: eight thousand seven hundred thirty five hypotheses over the clinical timelines of vaccine adverse-event reports. Extensive experiments with five model configurations show that systems fail mostly at choosing which records count, a verdict-evidence gap that accuracy alone conceals.

Sources in this cluster