Impact of Language Guidance: A Reproducibility Study

Cherish Puniani [email protected], Advika Sinha [email protected], Shree Singhi [email protected], Aayan Yadav [email protected] · 2025

Read Impact of Language Guidance: A Reproducibility Study by Cherish Puniani [email protected], Advika Sinha [email protected], Shree Singhi [email protected], Aayan Yadav [email protected] (2025) 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-2504-08140.txt.

Description

Four students at an Indian institute of technology, Cherish Puniani, Advika Sinha, Shree Singhi, and Aayan Yadav, repeat an earlier study on teaching computers from images and text in a 2025 paper. The earlier work claimed that words help a model pick better pairs of images to learn from.

The team reran the tests and found a flaw: the captions in the large data set used were of low quality. Swapping in new captions from an existing caption-writing model improved the results. They also built a new way to grade how well such models grasp meaning, based on tools that show what a model looks at. Their paper is a check on a published claim, and it shows why data quality matters.

Catalog record

CREATORCherish Puniani [email protected], Advika Sinha [email protected], Shree Singhi [email protected], Aayan Yadav [email protected]
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-08140

Browse the full library · About the Tower of Babel