# Pingala: Prosody-Aware Decoding for Sanskrit Poetry Generation

arXiv: https://arxiv.org/abs/2603.24413 · HTML: https://mlsanity.com/paper/2603.24413/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 28 references that are in the map.
- First posted on arXiv 25 Mar 2026.

## Builds on (most cited first, 23 of 23)

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [SimCSE: Simple Contrastive Learning of Sentence Embeddings](https://mlsanity.com/paper/2104.08821/) (2021)
- [BLEURT: Learning Robust Metrics for Text Generation](https://mlsanity.com/paper/2004.04696/) (2020)
- [COMET: A Neural Framework for MT Evaluation](https://mlsanity.com/paper/2009.09025/) (2020)
- [Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation](https://mlsanity.com/paper/2004.09813/) (2020)
- [Representation Degeneration Problem in Training Natural Language Generation Models](https://mlsanity.com/paper/1907.12009/) (2019)
- [InfoXLM: An Information-Theoretic Framework for Cross-Lingual Language Model Pre-Training](https://mlsanity.com/paper/2007.07834/) (2020)
- [Fine-grained Analysis of Sentence Embeddings Using Auxiliary Prediction Tasks](https://mlsanity.com/paper/1608.04207/) (2016)
- [Poly-encoders: Transformer Architectures and Pre-training Strategies for Fast and Accurate Multi-sentence Scoring](https://mlsanity.com/paper/1905.01969/) (2019)
- [Tokenization Impacts Multilingual Language Modeling: Assessing Vocabulary Allocation and Overlap Across Languages](https://mlsanity.com/paper/2305.17179/) (2023)
- [TransQuest: Translation Quality Estimation with Cross-lingual Transformers](https://mlsanity.com/paper/2011.01536/) (2020)
- [Semantic Answer Similarity for Evaluating Question Answering Models](https://mlsanity.com/paper/2108.06130/) (2021)
- [An Isotropy Analysis in the Multilingual BERT Embedding Space](https://mlsanity.com/paper/2110.04504/) (2021)
- [Translation Error Detection as Rationale Extraction](https://mlsanity.com/paper/2108.12197/) (2021)
- [In Defense of Cross-Encoders for Zero-Shot Retrieval](https://mlsanity.com/paper/2212.06121/) (2022)
- [Romanization-based Large-scale Adaptation of Multilingual Language Models](https://mlsanity.com/paper/2304.08865/) (2023)
- [Automatic Speech Recognition in Sanskrit: A New Speech Corpus and Modelling Insights](https://mlsanity.com/paper/2106.05852/) (2021)
- [Fine-Grained Reward Optimization for Machine Translation using Error Severity Mappings](https://mlsanity.com/paper/2411.05986/) (2024)
- [Chandomitra: Towards Generating Structured Sanskrit Poetry from Natural Language Inputs](https://mlsanity.com/paper/2506.00815/) (2025)
- [The Devil is in the Details: On Models and Training Regimes for Few-Shot Intent Classification](https://mlsanity.com/paper/2210.06440/) (2022)
- [Mitrasamgraha: A Comprehensive Classical Sanskrit Machine Translation Dataset](https://mlsanity.com/paper/2601.07314/) (2026)
- [Calibration Meets Reality: Making Machine Learning Predictions Trustworthy](https://mlsanity.com/paper/2509.23665/) (2025)

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

- [Padamitra: Grounded Glossary Generation for Classical Sanskrit](https://mlsanity.com/paper/2608.25038/) (2026)

## Abstract (arXiv)

Poetry generation in Sanskrit typically requires the verse to be semantically coherent and adhere to strict prosodic rules. In Sanskrit prosody, every line of a verse is typically a fixed length sequence of syllables adhering to prescribed binary patterns of syllable weights. We observe that instead of treating a verse as a monolithic sequence, segmenting them as grouped-lines leads to significant improvement in semantic coherence by 10% with comparable metrical adherence. Specifically, Pingala, our proposed decoding approach is designed to encourage every line to have well-formed words and our token selection biases the model towards it by preferring longer tokens. Writing in Sanskrit follows phonemic orthography, hence using a phonetically aware transliteration scheme, SLP1, increased the metrical alignment by 46% with comparable semantic similarity, for a instruction fine-tuned large language models like Phi-4. We also introduce a new approach for reference-free evaluation using cross-encoders which achieved better alignment with true poetry instances.