NoLima

NoLiMa: Long-Context Evaluation Beyond Literal Matching This repository contains the data associated with our ICML 2025 paper, "NoLiMa: Long-Context Evaluation Beyond Literal Matching". Abstract Recent large language models (LLMs) support long contexts ranging from 128K to 1M tokens. A popular method for evaluating the

Papers using NoLima (3)

Logit-Contribution Scoring Identifies Non-Literal Retrieval Heads2026NoLiMa: Long-Context Evaluation Beyond Literal Matching2025 · 4 citesTagging-Augmented Generation: Assisting Language Models in Finding Intricate Knowledge In Long Contexts2025