Why Stop at One Error? Benchmarking LLMs as Data Science Code Debuggers for Multi-Hop and Multi-Bug Errors

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Yang, Zhiyu, Wang, Shuo, Yan, Yukun, Deng, Yang
Format:	Preprint
Published:	2025
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2503.22388