Long Input Benchmark for Russian Analysis

TL;DR

This paper introduces LIBRA, a comprehensive benchmark with datasets and tools to evaluate the ability of language models to understand long Russian texts across various complexity levels and context lengths.

Contribution

The paper presents LIBRA, a new long-input benchmark for Russian NLP, including datasets, code, and a leaderboard to facilitate research on long-context understanding.

Findings

LIBRA covers 21 datasets across four complexity groups.

Models are evaluated on context lengths from 4k to 128k tokens.

Open-source resources are provided for community use.

Abstract

Recent advancements in Natural Language Processing (NLP) have fostered the development of Large Language Models (LLMs) that can solve an immense variety of tasks. One of the key aspects of their application is their ability to work with long text documents and to process long sequences of tokens. This has created a demand for proper evaluation of long-context understanding. To address this need for the Russian language, we propose LIBRA (Long Input Benchmark for Russian Analysis), which comprises 21 adapted datasets to study the LLM's abilities to understand long texts thoroughly. The tests are divided into four complexity groups and allow the evaluation of models across various context lengths ranging from 4k up to 128k tokens. We provide the open-source datasets, codebase, and public leaderboard for LIBRA to guide forthcoming research.