BAMBOO: A Comprehensive Benchmark for Evaluating Long Text Modeling Capacities of Large Language Models
This addresses the need for standardized evaluation of long context abilities in LLMs for researchers and developers, though it is incremental as it builds on existing benchmarking efforts.
The authors tackled the problem of evaluating long text modeling capabilities in large language models by introducing BAMBOO, a multi-task benchmark covering 10 datasets from 5 tasks, and found that it provides a comprehensive assessment with experiments on five models.
Large language models (LLMs) have achieved dramatic proficiency over NLP tasks with normal length. Recently, multiple studies have committed to extending the context length and enhancing the long text modeling capabilities of LLMs. To comprehensively evaluate the long context ability of LLMs, we propose BAMBOO, a multi-task long context benchmark. BAMBOO has been designed with four principles: comprehensive capacity evaluation, avoidance of data contamination, accurate automatic evaluation, and different length levels. It consists of 10 datasets from 5 different long text understanding tasks, i.e. question answering, hallucination detection, text sorting, language modeling, and code completion, to cover core capacities and various domains of LLMs. We conduct experiments with five long context models on BAMBOO and further discuss four key research questions of long text. We also qualitatively analyze current long context models and point out future directions for enhancing long text modeling capacities. We release our data, prompts, and code at https://github.com/RUCAIBox/BAMBOO.