# Learning Moderately Input-Sensitive Functions: A Case Study in QR Code Decoding

Source: [Learning Moderately Input-Sensitive Functions: A Case Study in QR Code Decoding](<https://www.sota2.com/research/paper/68fa799f1e209131cea7e7ee>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T05:01:45.766Z

Authors: Kazuki Yoda, Kazuhiko Kawamoto, Hiroshi Kera

Published: 2025-06-25T10:37:39.000Z

## Abstract

The hardness of learning a function that attains a target task relates to its input-sensitivity. For example, image classification tasks are input-insensitive as minor corruptions should not affect the classification results, whereas arithmetic and symbolic computation, which have been recently attracting interest, are highly input-sensitive as each input variable connects to the computation results. This study presents the first learning-based Quick Response (QR) code decoding and investigates learning functions of medium sensitivity. Our experiments reveal that Transformers can successfully decode QR codes, even beyond the theoretical error-correction limit, by learning the structure of embedded texts. They generalize from English-rich training data to other languages and even random strings. Moreover, we observe that the Transformer-based QR decoder focuses on data bits while ignoring error-correction bits, suggesting a decoding mechanism distinct from standard QR code readers.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2506.20305>)

[PDF](<https://arxiv.org/pdf/2506.20305.pdf>)