# A Case Study of Cross-Lingual Zero-Shot Generalization for Classical Languages in LLMs

arXiv: https://arxiv.org/abs/2505.13173 · HTML: https://mlsanity.com/paper/2505.13173/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 54 references that are in the map.
- First posted on arXiv 19 May 2025.

## Builds on (most cited first, 24 of 24)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [Enriching Word Vectors with Subword Information](https://mlsanity.com/paper/1607.04606/) (2016)
- [Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks](https://mlsanity.com/paper/2204.07705/) (2022)
- [Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph](https://mlsanity.com/paper/2307.07697/) (2023)
- [When Being Unseen from mBERT is just the Beginning: Handling New Languages With Multilingual Language Models](https://mlsanity.com/paper/2010.12858/) (2020)
- [LLMs Are Few-Shot In-Context Low-Resource Language Learners](https://mlsanity.com/paper/2403.16512/) (2024)
- [Generate-on-Graph: Treat LLM as both Agent and KG in Incomplete Knowledge Graph Question Answering](https://mlsanity.com/paper/2404.14741/) (2024)
- [Match the Script, Adapt if Multilingual: Analyzing the Effect of Multilingual Pretraining on Cross-lingual Transferability](https://mlsanity.com/paper/2203.10753/) (2022)
- [One Model is All You Need: ByT5-Sanskrit, a Unified Model for Sanskrit NLP Tasks](https://mlsanity.com/paper/2409.13920/) (2024)
- [Sangrahaka: A Tool for Annotating and Querying Knowledge Graphs](https://mlsanity.com/paper/2107.02782/) (2021)
- [Framework for Question-Answering in Sanskrit through Automated Construction of Knowledge Graphs](https://mlsanity.com/paper/2310.07848/) (2023)
- [Systematic Investigation of Strategies Tailored for Low-Resource Settings for Low-Resource Dependency Parsing](https://mlsanity.com/paper/2201.11374/) (2022)
- [S=amayik: A Benchmark and Dataset for English-Sanskrit Translation](https://mlsanity.com/paper/2305.14004/) (2023)
- [Semantic Annotation and Querying Framework based on Semi-structured Ayurvedic Text](https://mlsanity.com/paper/2202.00216/) (2022)
- [Deep Exploration of Cross-Lingual Zero-Shot Generalization in Instruction Tuning](https://mlsanity.com/paper/2406.08796/) (2024)
- [Sanskrit Knowledge-based Systems: Annotation and Computational Tools](https://mlsanity.com/paper/2406.18276/) (2024)

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

- [Semantic Alignment across Ancient Egyptian Language Stages via Normalization-Aware Multitask Learning](https://mlsanity.com/paper/2603.24258/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) have demonstrated remarkable generalization capabilities across diverse tasks and languages. In this study, we focus on natural language understanding in three classical languages -- Sanskrit, Ancient Greek and Latin -- to investigate the factors affecting cross-lingual zero-shot generalization. First, we explore named entity recognition and machine translation into English. While LLMs perform equal to or better than fine-tuned baselines on out-of-domain data, smaller models often struggle, especially with niche or abstract entity types. In addition, we concentrate on Sanskrit by presenting a factoid question-answering (QA) dataset and show that incorporating context via retrieval-augmented generation approach significantly boosts performance. In contrast, we observe pronounced performance drops for smaller LLMs across these QA tasks. These results suggest model scale as an important factor influencing cross-lingual generalization. Assuming that models used such as GPT-4o and Llama-3.1 are not instruction fine-tuned on classical languages, our findings provide insights into how LLMs may generalize on these languages and their consequent utility in classical studies.