AutoResearchBench: Benchmarking AI Agents on Complex Scientific Literature Discovery

TL;DR

AutoResearchBench is a new benchmark designed to evaluate AI agents' ability to autonomously discover and collect scientific literature, emphasizing in-depth understanding and open-ended search tasks.

Contribution

The paper introduces AutoResearchBench, a research-oriented, literature-focused benchmark with two complex tasks, and provides the dataset, evaluation pipeline, and code for future research.

Findings

State-of-the-art LLMs achieve only around 9.4% accuracy on Deep Research.

Most baseline models perform below 5% on the benchmark.

The benchmark is highly challenging compared to previous web-browsing benchmarks.

Abstract

Autonomous scientific research is significantly advanced thanks to the development of AI agents. One key step in this process is finding the right scientific literature, whether to explore existing knowledge for a research problem, or to acquire evidence for verifying assumptions and supporting claims. To assess AI agents' capability in driving this process, we present AutoResearchBench, a dedicated benchmark for autonomous scientific literature discovery. AutoResearchBench consists of two complementary task types: (1) Deep Research, which requires tracking down a specific target paper through a progressive, multi-step probing process, and (2) Wide Research, which requires comprehensively collecting a set of papers satisfying given conditions. Compared to previous benchmarks on agentic web browsing, AutoResearchBench is distinguished along three dimensions: it is research-oriented, calling for in-depth comprehension of scientific concepts; literature-focused, demanding fine-grained utilization of detailed information; and open-ended, involving an unknown number of qualified papers and thus requiring deliberate reasoning and search throughout. These properties make AutoResearchBench uniquely suited for evaluating autonomous research capabilities, and extraordinarily challenging. Even the most powerful LLMs, despite having largely conquered general agentic web-browsing benchmarks such as BrowseComp, achieve only 9.39% accuracy on Deep Research and 9.31% IoU on Wide Research, while many other strong baselines fall below 5%. We publicly release the dataset and evaluation pipeline to facilitate future research in this direction. We publicly release the dataset, evaluation pipeline, and code at https://github.com/CherYou/AutoResearchBench.