ESG-Bench: Benchmarking Long-Context ESG Reports for Hallucination Mitigation

Siqi Sun, Ben Peng Wu, Mali Jin, Peizhen Bai, Hanpei Zhang, Xingyi Song · 2026

Read ESG-Bench: Benchmarking Long-Context ESG Reports for Hallucination Mitigation by Siqi Sun, Ben Peng Wu, Mali Jin, Peizhen Bai, Hanpei Zhang, Xingyi Song (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-13154.txt.

Description

Siqi Sun and co-authors introduce ESG-Bench, a benchmark for reading long ESG reports without hallucination. Corporate ESG reports are now legal duties in many places, yet they run long and complex. The bench holds human-labeled question-answer pairs tied to real report text. Fine labels mark whether an answer is backed by the source or made up. Task-specific chain-of-thought prompts and tuned models cut false claims a lot. Gains also carried to standard question answering sets.

Catalog record

CREATORSiqi Sun, Ben Peng Wu, Mali Jin, Peizhen Bai, Hanpei Zhang, Xingyi Song
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE34189
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.13154

Browse the full library · About the Tower of Babel