作者位次 #2/3ACL 2026Poster· Natural Language Processing
When Benchmarks Leak: Inference-Time Decontamination for LLM s
Jianzhe Chai · YU Zhe · Jun Sakuma
摘要
Benchmark-based evaluation is the de facto standard for comparing large language models (LLMs). However, its reliability is increasingly threatened by test set contamination, where test samples or their close variants leak into training data and artificially inflate reported performance. To addre...