#artificial intelligence
Jun 2026
DiagFlowBench: Evaluating How Language Models Handle Off-Procedure Inputs in Grounded Diagnostic Dialogue
DiagFlowBench is introduced, a dataset of 50 industrial diagnostic flowcharts from a consumer manufacturer converted into 1,676 multi-turn conversations that contrast compliant with off-procedure utterances that reveals high variability in abstention rates.
Guillermo Gil de Avalle, L. Maruster, Shaina Raza et al.
· arXiv.org · 0 citations