repository icon
microsoft / promptbench Project repository

A unified evaluation framework for large language models

Submit a report

FIRST INTERACTION

WITHIN61 DAYS

REVIEW

WITHIN61 DAYS

FIX

WITHINN/A DAYS

Arbitrary Code Execution via Unsafe eval() in promptbench.MMMU Dataset Loader
cyfra07
duplicate
High
Arbitrary Code Execution via Unchecked eval() in PromptBench (MMMU Dataset)
eladbeber
informative
High