作者位次 #2/11ACL 2026Poster· Natural Language Processing
Test of Time: Rethinking Temporal Signal of Benchmark Contamination
Terry Jingchen Zhang · Gopal Dev · Ning Wang · Max Obreiter · Punya Syon Pandey · Keenan Samway · Wenyuan Jiang · Yinya Huang · Bernhard Schölkopf · Mrinmaya Sachan · Zhijing Jin
摘要
Post-cutoff performance decay of LLMs has been widely interpreted as a temporal signal for benchmark contamination, where public information released before the training cutoff may have been included into training corpora and inflated model performance by memorization. We critically examine this...