AI Insight
Researchers have developed ReplicatorBench, a new benchmark for evaluating AI agents' ability to replicate scientific studies in social and behavioral sciences. Unlike existing benchmarks that only test reproduction with available code and data, ReplicatorBench includes both replicable and non-replicable research claims and evaluates agents across three stages: data retrieval, experiment execution, and results interpretation. Testing with ReplicatorAgent across four large language models revealed that while AI agents can effectively design and execute experiments, they struggle significantly with retrieving necessary resources like new datasets needed for replication.
Why it matters
This benchmark addresses a critical gap in scientific validation by testing whether AI can assist in the replication crisis affecting social and behavioral sciences. The findings suggest that current AI systems could potentially automate parts of the scientific review process, though significant limitations in data retrieval must be overcome before they can reliably assess research replicability.
Understand the Science
arXiv:2602.11354v3 Announce Type: replace
Abstract: The literature has witnessed an emerging interest in AI agents for automated assessment of scientific papers. Existing benchmarks focus primarily on the computational aspect of this task, testing agents’ ability to reproduce or replicate research outcomes when having access to the code and data. This setting, while foundational, (1) fails to capture the inconsistent availability of new data for replication as opposed to reproduction, and (2) lacks ground-truth diversity by focusing only on reproducible papers, thereby failing to evaluate an agent’s ability to identify non-replicable research. Furthermore, most benchmarks only evaluate outcomes rather than the replication process. In response, we introduce ReplicatorBench, an end-to-end benchmark, including human-verified replicable and non-replicable research claims in social and behavioral sciences for evaluating AI agents in research replication across three stages: (1) extraction and retrieval of replication data; (2) design and execution of computational experiments; and (3) interpretation of results, allowing a test of AI agents’ capability to mimic the activities of human replicators in real world. To set a baseline of AI agents’ capability, we develop ReplicatorAgent, an agentic framework equipped with necessary tools, like web search and iterative interaction with sandboxed environments, to accomplish tasks in ReplicatorBench. We evaluate ReplicatorAgent across four underlying large language models (LLMs), as well as different design choices of programming language and levels of code access. Our findings reveal that while current LLM agents are capable of effectively designing and executing computational experiments, they struggle with retrieving resources, such as new data, necessary to replicate a claim. All code and data are publicly available at https://github.com/CenterForOpenScience/llm-benchmarking.
Source: ReplicatorBench: Benchmarking LLM Agents for Replicability in Social and Behavioral Sciences