InterveneBench: Benchmarking LLMs for Intervention Reasoning and Causal Study Design in Real Social Systems

TL;DR

InterveneBench is a new benchmark for evaluating large language models' ability to reason about intervention and causal inference in realistic social science scenarios, revealing current limitations and proposing a multi-agent solution.

Contribution

The paper introduces InterveneBench, a benchmark based on real social science studies, and proposes STRIDES, a multi-agent framework that improves LLM reasoning in causal intervention tasks.

Findings

State-of-the-art LLMs perform poorly on InterveneBench.

STRIDES significantly outperforms existing reasoning models.

InterveneBench covers 744 diverse social science studies.

Abstract

Causal inference in social science relies on end-to-end, intervention-centered research-design reasoning grounded in real-world policy interventions, but current benchmarks fail to evaluate this capability of large language models (LLMs). We present InterveneBench, a benchmark designed to assess such reasoning in realistic social settings. Each instance in InterveneBench is derived from an empirical social science study and requires models to reason about policy interventions and identification assumptions without access to predefined causal graphs or structural equations. InterveneBench comprises 744 peer-reviewed studies across diverse policy domains. Experimental results show that state-of-the-art LLMs struggle under this setting. To address this limitation, we further propose a multi-agent framework, STRIDES. It achieves significant performance improvements over state-of-the-art reasoning models. Our code and data are available at https://github.com/Sii-yuning/STRIDES.