The authors define contamination resistance as a benchmark property and show that most tested LLMs score near zero on Caesar-cipher encoding and decoding when the shift is not 3 and the text is random nonsense.
Leak, cheat, repeat: Data contamination and evaluation malpractices in closed-source llms
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Towards Contamination Resistant Benchmarks
The authors define contamination resistance as a benchmark property and show that most tested LLMs score near zero on Caesar-cipher encoding and decoding when the shift is not 3 and the text is random nonsense.