Zenodo DOI Release for “Rewarding the Scientific Process”

This release is created for the purpose of obtaining a Zenodo DOI for the paper “Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis.”

The code associated with this paper is located in the dataprm/ directory of this repository. This directory contains the implementation and related resources for DataPRM, including the training and application code for process-level reward modeling in agentic data analysis.

Please refer to the README and the files under dataprm/ for usage instructions, dependencies, and additional details.