Holographic Transformers for Complex-Valued Signal Processing: Integrating Phase Interference into Self-Attention
Enhao Huang 1 , Zhiyu Zhang 1 , Tianxiang Xu 2 , Chunshu Xia 1 , Kaichun Hu 1 Yuchen Yang 1 , Tongtong Pan 1 , Dong Dong 1 , Zhan Qin 1 \sthanks Corresponding author. · 2025
Read Holographic Transformers for Complex-Valued Signal Processing: Integrating Phase Interference into Self-Attention by Enhao Huang 1 , Zhiyu Zhang 1 , Tianxiang Xu 2 , Chunshu Xia 1 , Kaichun Hu 1 Yuchen Yang 1 , Tongtong Pan 1 , Dong Dong 1 , Zhan Qin 1 \sthanks Corresponding author. (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-2509-19331.txt.
Description
Signals like radar returns and wireless channels carry both strength and phase, but most neural networks attend only to real-valued similarity. Enhao Huang and eight co-authors build a Holographic Transformer that bakes wave interference into the attention mechanism itself. In their design, interactions between signal parts are modulated by relative phase, and values combine coherently, the way light waves add. A two-headed decoder rebuilds the input while predicting the task output, which stops the model from letting phase collapse.
The authors show their attention acts as a discrete interference operator and stays phase-consistent when signals mix. Experiments cover polarimetric radar image classification and wireless channel prediction. Scores were strong on both, with low regression error and extra robustness when phases were perturbed. The work offers a physics-based framework for learning from complex-valued signals as coherent wholes.
Catalog record
| CREATOR | Enhao Huang 1 , Zhiyu Zhang 1 , Tianxiang Xu 2 , Chunshu Xia 1 , Kaichun Hu 1 Yuchen Yang 1 , Tongtong Pan 1 , Dong Dong 1 , Zhan Qin 1 \sthanks Corresponding author. |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-19331