Hierarchic-EEG2Text: Assessing EEG-To-Text Decoding across Hierarchical Abstraction Levels CCS: Applied computing Bioinformatics CCS: Computing methodologies Multi-task learning

Anupam Sharma, Harish Katti, Prajwal Singh, Shanmuganathan Raman, Krishna Miyapuram · 2026

Read Hierarchic-EEG2Text: Assessing EEG-To-Text Decoding across Hierarchical Abstraction Levels CCS: Applied computing Bioinformatics CCS: Computing methodologies Multi-task learning by Anupam Sharma, Harish Katti, Prajwal Singh, Shanmuganathan Raman, Krishna Miyapuram (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-2602-20932.txt.

Description

Brain waves recorded from the scalp are noisy, so decoding fine object detail from them is hard, but broader categories may still show through. Anupam Sharma and four co-authors test that idea in this 2026 arXiv paper, asking whether EEG captures object representations at several levels of abstraction. They evaluate models across related classification tasks called episodes, sampled with awareness of the WordNet hierarchy.

Their framework is the largest episodic EEG study of its kind. It covers nearly a million EEG samples under more than sixteen hundred object labels, from over two hundred sixty participants in the PEERS dataset. Models do better when categories come from higher levels of the hierarchy, showing sensitivity to abstraction. Abstraction depth emerges as an underexplored dimension of EEG decoding.

Catalog record

CREATORAnupam Sharma, Harish Katti, Prajwal Singh, Shanmuganathan Raman, Krishna Miyapuram
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-20932

Browse the full library · About the Tower of Babel