Reading Why: Born-Measured Attention with Exact Word Evidence on Light and Qubits
Abstract
When a language model answers, we can usually only estimate why. We build a transformer in which every step is a Born measurement - attention |⟨ q|k⟩|4 normalised over each row, a measured product in place of the feed-forward nonlinearity, and a Born readout. With its denominators held at their values for the input, each class probability is a polynomial in the word amplitudes, so its integrated-gradient evidence is computed by a finite quadrature with no quadrature error: the words' contributions sum to the class probability above its no-information share, to 10−11 or better. On e-SNLI natural language inference, with about 98k parameters outside the embedding, the model reaches 76.98 ± 0.17% test accuracy, level with a softmax transformer of the same size (76.62 ± 0.36%). Deleting the word its evidence ranks first changes the answer in 45% of test pairs, out of the 53% any single deletion can change, against 32% for the softmax model explained the same way and 40% by its attention x gradient. Attention can look more like the words annotators highlighted, yet accounts for the answer less well: deleting the word it ranks first changes the Born Transformer's answer in 32% of pairs. The same weights compile exactly to optoelectronic optics and to qubits, and a coherent attention layer becomes one circuit equal to the model (852 CNOTs on 23 qubits). On IBM's ibm_kingston, reading only the attention overlaps by Hadamard tests, the device gave the model's answers on two short test pairs and kept or changed them under deletion as the model does; the evidence itself did not yet survive the device, and we work out the shots and calibration that measuring it on hardware would need.
Related
- A Quantum Roadmap for Softmax Attention: Exact Born-Rule Analogs for Softmax Attention on the Probability Simplex
- Standard estimators cannot represent fault-tolerant workloads at measured error rates: evaluated, evidence-based uncertainty for quantum resource estimation
- Seeing Why: Exact Evidence Maps for Quantum and Optical Image Classifiers
- Spin Kerr-cat qubits
- Entanglement without the Entangled State: A Local Theory of Qubits with an Extended Born Rule
BibTeX
@misc{tongloy2026reading,
title = {{Reading Why: Born-Measured Attention with Exact Word Evidence on Light and Qubits}},
author = {Teerawat Tongloy and Jiraporn Pongsopa and Rangsan Jomtarak},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.05355},
eprint = {2611.05355},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.05355},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Reading Why: Born-Measured Attention with Exact Word Evidence on Light and Qubits
AU - Teerawat Tongloy
AU - Jiraporn Pongsopa
AU - Rangsan Jomtarak
PY - 2026
DA - 2026/11/03
PB - arXiv
DO - 10.48550/arxiv.2611.05355
UR - https://arxiv.org/abs/2611.05355
AB - When a language model answers, we can usually only estimate why. We build a transformer in which every step is a Born measurement - attention |⟨ q|k⟩|^4 normalised over each row, a measured product in place of the feed-forward nonlinearity, and a Born readout. With its denominators held at their values for the input, each class probability is a polynomial in the word amplitudes, so its integrated-gradient evidence is computed by a finite quadrature with no quadrature error: the words' contributions sum to the class probability above its no-information share, to 10^−11 or better. On e-SNLI natural language inference, with about 98k parameters outside the embedding, the model reaches 76.98 ± 0.17% test accuracy, level with a softmax transformer of the same size (76.62 ± 0.36%). Deleting the word its evidence ranks first changes the answer in 45% of test pairs, out of the 53% any single deletion can change, against 32% for the softmax model explained the same way and 40% by its attention x gradient. Attention can look more like the words annotators highlighted, yet accounts for the answer less well: deleting the word it ranks first changes the Born Transformer's answer in 32% of pairs. The same weights compile exactly to optoelectronic optics and to qubits, and a coherent attention layer becomes one circuit equal to the model (852 CNOTs on 23 qubits). On IBM's ibm_kingston, reading only the attention overlaps by Hadamard tests, the device gave the model's answers on two short test pairs and kept or changed them under deletion as the model does; the evidence itself did not yet survive the device, and we work out the shots and calibration that measuring it on hardware would need.
ER -