# I'm Sorry, but I Can't Help with Braille: Revealing Accessibility Failures in State-of-the-Art LLMs

arXiv: https://arxiv.org/abs/2607.11893 · HTML: https://mlsanity.com/paper/2607.11893/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 15 references that are in the map.
- First posted on arXiv 7 May 2026.

## Builds on (most cited first, 6 of 6)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [CIDEr: Consensus-based Image Description Evaluation](https://mlsanity.com/paper/1411.5726/) (2014)
- [A Call for Clarity in Reporting BLEU Scores](https://mlsanity.com/paper/1804.08771/) (2018)
- [The State and Fate of Linguistic Diversity and Inclusion in the NLP World](https://mlsanity.com/paper/2004.09095/) (2020)
- [BrailleLLM: Braille Instruction Tuning with Large Language Models for Braille Domain Tasks](https://mlsanity.com/paper/2510.18288/) (2025)

## Built on by (most cited first, 1 of 1)

- [BrailleBench: Investigating Multi-Criteria Braille Comprehension in Large Language Models](https://mlsanity.com/paper/2608.27268/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) perform strongly on many language tasks, but their capability in structurally constrained, accessibility-critical modalities such as Braille remains unclear. We evaluate state-of-the-art LLMs on bidirectional Korean-Braille translation using a human-annotated dataset. Despite expectations that multilingual, instruction-tuned models can generalize to Braille via text representations, we find consistently poor, unstable outputs and substantial disagreement with human judgments. These results point to missing Braille-aware tokenization and weak alignment between Korean and Braille patterns. In contrast, supervised fine-tuning of a small model (T5-small) on the same data yields large and stable gains over zero-shot and prompted LLM baselines across standard metrics (SacreBLEU, ChrF++, CER, BLEU, ROUGE-L, METEOR, CIDEr). Our findings reveal a systematic limitation of current LLMs and demonstrate the effectiveness of modest task-specific supervision.