A constrained-prompt LLM credit-risk explainer inverted the sign of three of four supplied drivers in an audited case, while the stacking ensemble's AUC gain over random forest was real but operationally small.
‘Why should I trust you?’ Ex- plaining the predictions of any classifier,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Accurate Ensembles, Fragile Narratives: Multi-Scale Stacking and a Fidelity Audit of LLM-Generated Explanations for Credit Risk
A constrained-prompt LLM credit-risk explainer inverted the sign of three of four supplied drivers in an audited case, while the stacking ensemble's AUC gain over random forest was real but operationally small.