Deep Sequence Modeling with Quantum Dynamics: Language as a Wave Function
Ahmed Nebli, Hadi Saadatdoorabi, Kevin Yam · 2026
Read Deep Sequence Modeling with Quantum Dynamics: Language as a Wave Function by Ahmed Nebli, Hadi Saadatdoorabi, Kevin Yam (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-2602-22255.txt.
Description
Most sequence models juggle competing readings with gating, suppressing losers to settle on one interpretation. Ahmed Nebli and two co-authors try a quantum route instead: the model's latent state is a complex wave function evolving under a learned, time-dependent Hamiltonian. Quantum interference then cancels conflicting readings while reinforcing compatible ones, with no gates required.
The dynamics stay strictly unitary, preserving the state's norm exactly at every step through a Cayley discretization, and token probabilities come from the Born rule. The paper's main theorem separates this readout's power from classical rivals on a family of disambiguation tasks. Language modeling thus gains a wave-based alternative whose advantage is proven, not just promised.
Catalog record
| CREATOR | Ahmed Nebli, Hadi Saadatdoorabi, Kevin Yam |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 178 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-22255