SciDER: Scientific Data-centric End-to-end Researcher

TL;DR

SciDER is a novel data-centric system that automates the entire scientific research process, from data analysis to hypothesis generation and experimentation, enhancing autonomous scientific discovery.

Contribution

The paper introduces SciDER, a specialized, modular system that improves scientific data processing and hypothesis generation over existing general-purpose agents.

Findings

SciDER outperforms state-of-the-art models on three scientific discovery benchmarks.

It features a self-evolving memory and critic-led feedback loop for continuous improvement.

Distributed as a Python package with a web interface for accessibility.

Abstract

Automated scientific discovery with large language models is transforming the research lifecycle from ideation to experimentation, yet existing agents struggle to autonomously process raw data collected from scientific experiments. We introduce SciDER, a data-centric end-to-end system that automates the research lifecycle. Unlike traditional frameworks, our specialized agents collaboratively parse and analyze raw scientific data, generate hypotheses and experimental designs grounded in specific data characteristics, and write and execute corresponding code. Evaluation on three benchmarks shows SciDER excels in specialized data-driven scientific discovery and outperforms general-purpose agents and state-of-the-art models through its self-evolving memory and critic-led feedback loop. Distributed as a modular Python package, we also provide easy-to-use PyPI packages with a lightweight web interface to accelerate autonomous, data-driven research and aim to be accessible to all researchers and developers.