Gold-Guided Programmatic Distillation Boosts Financial Reasoning in Smaller LLMs

Jul 17, 2026

Researchers introduce a method that distills numerical reasoning abilities from a large language model into a smaller one using execution-verified Python programs, rather than natural-language rationales. On the TAT-QA benchmark, their 7B-parameter student model achieves 87.00 EM / 87.18 F1, outperforming both its 72B-parameter teacher (78.46 EM) and strong existing baselines. The approach includes an iterative recovery stage to further improve training by incorporating newly verified programs.

Why it matters: This work demonstrates that programmatic distillation can enable smaller models to outperform much larger ones in complex financial reasoning tasks, potentially reducing computational costs for high-accuracy applications.

Full story at: arXiv Computation and Language ↗