# Semantic Alignment across Ancient Egyptian Language Stages via Normalization-Aware Multitask Learning

arXiv: https://arxiv.org/abs/2603.24258 · HTML: https://mlsanity.com/paper/2603.24258/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 27 references that are in the map.
- First posted on arXiv 25 Mar 2026.

## Builds on (most cited first, 7 of 7)

- [Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics](https://mlsanity.com/paper/1705.07115/) (2017)
- [Cross-lingual Language Model Pretraining](https://mlsanity.com/paper/1901.07291/) (2019)
- [A Survey Of Cross-lingual Word Embedding Models](https://mlsanity.com/paper/1706.04902/) (2017)
- [Cross-lingual Alignment vs Joint Training: A Comparative Study and A Simple Unified Framework](https://mlsanity.com/paper/1910.04708/) (2019)
- [A Collaborative Ecosystem for Digital Coptic Studies](https://mlsanity.com/paper/1912.05082/) (2019)
- [A UD Treebank for Bohairic Coptic](https://mlsanity.com/paper/2504.18386/) (2025)
- [A Case Study of Cross-Lingual Zero-Shot Generalization for Classical Languages in LLMs](https://mlsanity.com/paper/2505.13173/) (2025)

## Abstract (arXiv)

We study word-level semantic alignment across four historical stages of Ancient Egyptian. These stages differ in script and orthography, and parallel data are scarce. We jointly train a compact encoder-decoder model with a shared byte-level tokenizer on all four stages, combining masked language modeling (MLM), translation language modeling (TLM), sequence-to-sequence translation, and part-of-speech tagging under a task-aware loss with fixed weights and uncertainty-based scaling. To reduce surface divergence we add Latin transliteration and IPA reconstruction as auxiliary views. We integrate these views through KL-based consistency and through embedding-level fusion. We evaluate alignment quality using pairwise metrics, specifically ROC-AUC and triplet accuracy, on curated Egyptian-English and intra-Egyptian cognate datasets. Translation yields the strongest gains. IPA with KL consistency improves cross-branch alignment, while early fusion demonstrates limited efficacy. Although the overall alignment remains limited, the findings provide a reproducible baseline and practical guidance for modeling historical languages under real constraints. They also show how normalization and task design shape what counts as alignment in typologically distant settings.