Relay-Bench: Evaluating LLMs on Multi-Domain Reasoning Chains

Relay-Bench：评估大语言模型在多领域推理链上的表现

Abstract: Introducing Relay-Bench, an unsaturated, holistic, text-only benchmark that measures LLMs’ ability to complete an assortment of tasks from distinct domains in a single prompt. The leading model, GPT-5.5 (xHigh), scores 43.3%.

摘要： 本文介绍了 Relay-Bench，这是一个非饱和、整体性、纯文本的基准测试，旨在衡量大语言模型（LLM）在单个提示词中完成不同领域任务组合的能力。目前领先的模型 GPT-5.5 (xHigh) 得分为 43.3%。

The test set entirely consists of composite problems: groups of single-domain subproblems that are strung together into challenges that require reasoning across multiple domains in combination. Many of these problems then have layers of complexity added through prompt encoding and deliberate context bloat.

测试集完全由复合问题组成：将一组单领域子问题串联起来，形成需要跨多个领域进行综合推理的挑战。其中许多问题还通过提示词编码和刻意的上下文冗余增加了复杂度。

Domains tested include visual reasoning, coding, math, information extraction (with a focus on web search), problem-solving, general knowledge, and data analysis. No restrictions are imposed outside of the model harness, and models are explicitly encouraged to leverage code-execution, web searches, and all available tools. All problems are composed of two to thirteen subproblems and do not require multi-modal input or output.

测试领域包括视觉推理、编程、数学、信息提取（侧重于网络搜索）、问题解决、常识和数据分析。除了模型框架本身的限制外，没有施加其他限制，并明确鼓励模型利用代码执行、网络搜索及所有可用工具。所有问题均由 2 到 13 个子问题组成，且不需要多模态输入或输出。