EIFBENCH: Extremely Complex Instruction Following Benchmark for Large Language Models

TL;DR

EIFBENCH is a new benchmark designed to evaluate large language models on complex, multi-task instructions with constraints, revealing significant performance gaps and highlighting the need for further optimization.

Contribution

The paper introduces EIFBENCH, a comprehensive benchmark for assessing LLMs on multi-task, constrained instructions, and proposes the SegPO algorithm to improve multi-task workflow execution.

Findings

Existing LLMs show large performance gaps on EIFBENCH.

EIFBENCH effectively simulates real-world complex scenarios.

SegPO enhances LLMs' multi-task handling capabilities.

Abstract

With the development and widespread application of large language models (LLMs), the new paradigm of "Model as Product" is rapidly evolving, and demands higher capabilities to address complex user needs, often requiring precise workflow execution which involves the accurate understanding of multiple tasks. However, existing benchmarks focusing on single-task environments with limited constraints lack the complexity required to fully reflect real-world scenarios. To bridge this gap, we present the Extremely Complex Instruction Following Benchmark (EIFBENCH), meticulously crafted to facilitate a more realistic and robust evaluation of LLMs. EIFBENCH not only includes multi-task scenarios that enable comprehensive assessment across diverse task types concurrently, but also integrates a variety of constraints, replicating complex operational environments. Furthermore, we propose the Segment Policy Optimization (SegPO) algorithm to enhance the LLM's ability to accurately fulfill multi-task workflow. Evaluations on EIFBENCH have unveiled considerable performance discrepancies in existing LLMs when challenged with these extremely complex instructions. This finding underscores the necessity for ongoing optimization to navigate the intricate challenges posed by LLM applications.