Inference-Time Decontamination: Reusing Leaked Benchmarks for Large Language Model Evaluation.
Qin ZhuQingyuan ChengRunyu PengXiaonan LiTengxiao LiuRu PengXipeng QiuXuanjing HuangPublished in: CoRR (2024)
Keyphrases
- language model
- language modeling
- n gram
- retrieval model
- document retrieval
- information retrieval
- query expansion
- ad hoc information retrieval
- language modelling
- probabilistic model
- language model for information retrieval
- test collection
- speech recognition
- query terms
- statistical language models
- smoothing methods
- context sensitive
- mixture model
- pseudo relevance feedback
- query specific
- translation model
- vector space model
- relevant documents
- document ranking
- bayesian networks