ConStat: Performance-Based Contamination Detection in Large Language Models
Public benchmarks play an essential role in the evaluation of large language models. However, data contamination can lead to inflated performance, rendering them unreliable for model comparison. It is therefore crucial to detect contamination and estimate its impact on measured performance. Unfortunately, existing detection methods can be easily evaded and fail to quantify contamination. To overcome these limitations, we propose a novel definition of contamination as artificially inflated and non-generalizing benchmark performance instead of the inclusion of benchmark samples in the training data. This perspective enables us to detect any model with inflated performance, i.e., performance that does not generalize to rephrased samples, synthetic samples from the same distribution, or different benchmarks for the same task. Based on this insight, we develop ConStat, a statistical method that reliably detects and quantifies contamination by comparing performance between a primary and reference benchmark relative to a set of reference models. We demonstrate the effectiveness of ConStat in an extensive evaluation of diverse model architectures, benchmarks, and contamination scenarios and find high levels of contamination in multiple popular models including Mistral, Llama, Yi, and the top-3 Open LLM Leaderboard models.
Code (0)
등록된 구현이 없습니다.
Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Towards Data Contamination Detection for Modern Large Language Models: Limitations, Inconsistencies, and Oracle Challenges
As large language models achieve increasingly impressive results, questions arise about whether such performance is from generalizability or mere data memorization. Thus, numerous data contamination detection methods hav…
MemorizationEvading Data Contamination Detection for Language Models is (too) Easy
Large language models are widespread, with their performance on benchmarks frequently guiding user preferences for one model over another. However, the vast amount of data these models are trained on can inadvertently le…
A Taxonomy for Data Contamination in Large Language Models
Large language models pretrained on extensive web corpora demonstrate remarkable performance across a wide range of downstream tasks. However, a growing concern is data contamination, where evaluation datasets may be con…
Question AnsweringLNE-Blocking: An Efficient Framework for Contamination Mitigation Evaluation on Large Language Models
The problem of data contamination is now almost inevitable during the development of large language models (LLMs), with the training data commonly integrating those evaluation benchmarks even unintentionally. This proble…
Generalization or Memorization: Data Contamination and Trustworthy Evaluation for Large Language Models
Recent statements about the impressive capabilities of large language models (LLMs) are usually supported by evaluating on open-access benchmarks. Considering the vast size and wide-ranging sources of LLMs' training data…
HumanEvalMemorization