1
All Leaks Count, Some Count More: Interpretable Temporal Contamination Detection and Mitigation in LLM Backtesting
时间污染让LLM回测结果虚高，这篇论文提出可解释的检测与缓解方法，让模型评估更可靠
arXiv:2602.17234v2 Announce Type: replace Abstract: Backtesting LLMs on resolved events assumes models reason only from pre-cutoff knowledge, yet pret…