New Gap-K% method detects LLM pretraining data

Researchers have developed a new method called Gap-K% to detect pretraining data used in large language models. This technique analyzes the gap between a model's top prediction and the actual target token, leveraging the gradient signals that are penalized during training. By incorporating local token correlations, Gap-K% significantly outperforms existing methods on benchmarks like WikiMIA and MIMIR, offering a more robust approach to identifying training data. AI

IMPACT Enhances transparency and accountability in LLM development by providing a tool to identify training data sources.

RANK_REASON The cluster contains an academic paper detailing a new method for detecting pretraining data in LLMs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Gap-K%
• Large Language Models
• MIMIR
• Minseo Kwak
• WikiMIA

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →