![[2602.20932] Hierarchic-EEG2Text: Assessing EEG-To-Text Decoding across Hierarchical Abstraction Levels](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.20932] Hierarchic-EEG2Text: Assessing EEG-To-Text Decoding across Hierarchical Abstraction Levels
Summary
This article presents a novel approach to EEG-to-text decoding, exploring how hierarchical abstraction levels affect classification performance using a large dataset of EEG samples.
Why It Matters
Understanding how EEG can decode complex cognitive processes at different abstraction levels has significant implications for advancements in brain-computer interfaces and cognitive neuroscience. This research could enhance our ability to interpret brain activity and improve applications in mental health and human-computer interaction.
Key Takeaways
- The study investigates EEG's ability to capture object representations across hierarchical levels.
- A new episodic analysis method is proposed, utilizing hierarchy-aware episode sampling from WordNet.
- The research presents the largest EEG dataset for text detection, enhancing understanding of neural dynamics.
- Higher abstraction levels in classification categories improve performance, indicating EEG sensitivity to abstraction depth.
- This work motivates further exploration of abstraction in EEG decoding for future research.
Computer Science > Machine Learning arXiv:2602.20932 (cs) [Submitted on 24 Feb 2026] Title:Hierarchic-EEG2Text: Assessing EEG-To-Text Decoding across Hierarchical Abstraction Levels Authors:Anupam Sharma, Harish Katti, Prajwal Singh, Shanmuganathan Raman, Krishna Miyapuram View a PDF of the paper titled Hierarchic-EEG2Text: Assessing EEG-To-Text Decoding across Hierarchical Abstraction Levels, by Anupam Sharma and 4 other authors View PDF HTML (experimental) Abstract:An electroencephalogram (EEG) records the spatially averaged electrical activity of neurons in the brain, measured from the human scalp. Prior studies have explored EEG-based classification of objects or concepts, often for passive viewing of briefly presented image or video stimuli, with limited classes. Because EEG exhibits a low signal-to-noise ratio, recognizing fine-grained representations across a large number of classes remains challenging; however, abstract-level object representations may exist. In this work, we investigate whether EEG captures object representations across multiple hierarchical levels, and propose episodic analysis, in which a Machine Learning (ML) model is evaluated across various, yet related, classification tasks (episodes). Unlike prior episodic EEG studies that rely on fixed or randomly sampled classes of equal cardinality, we adopt hierarchy-aware episode sampling using WordNet to generate episodes with variable classes of diverse hierarchy. We also present the largest episodic...