# Polite on the Surface, Broken in Practice: A Curated Dataset for Fixing Generation and Register Failures in Low-Resource Bangla Text Generation

arXiv: https://arxiv.org/abs/2605.22487 · HTML: https://mlsanity.com/paper/2605.22487/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 27 references that are in the map.
- First posted on arXiv 21 May 2026.

## Builds on (most cited first, 15 of 15)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [LIMA: Less Is More for Alignment](https://mlsanity.com/paper/2305.11206/) (2023)
- [BanglaBERT: Language Model Pretraining and Benchmarks for Low-Resource Language Understanding Evaluation in Bangla](https://mlsanity.com/paper/2101.00204/) (2021)
- [LLMs Are Few-Shot In-Context Low-Resource Language Learners](https://mlsanity.com/paper/2403.16512/) (2024)
- [Investigating the Translation Performance of a Large Multilingual Language Model: the Case of BLOOM](https://mlsanity.com/paper/2303.01911/) (2023)
- [A New Massive Multilingual Dataset for High-Performance Language Technologies](https://mlsanity.com/paper/2403.14009/) (2024)
- [Naamapadam: A Large-Scale Named Entity Annotated Data for Indic Languages](https://mlsanity.com/paper/2212.10168/) (2022)
- [Quality or Quantity? On Data Scale and Diversity in Adapting Large Language Models for Low-Resource Translation](https://mlsanity.com/paper/2408.12780/) (2024)
- [TituLLMs: A Family of Bangla LLMs with Comprehensive Benchmarking](https://mlsanity.com/paper/2502.11187/) (2025)
- [Chittron: An Automatic Bangla Image Captioning System](https://mlsanity.com/paper/1809.00339/) (2018)
- [Do Language Models Understand Honorific Systems in Javanese?](https://mlsanity.com/paper/2502.20864/) (2025)
- [Women, Infamous, and Exotic Beings: A Comparative Study of Honorific Usages in Wikipedia and LLMs for Bengali and Hindi](https://mlsanity.com/paper/2501.03479/) (2025)
- [Analyzing Syntactic Generalization Capacity of Pre-trained Language Models on Japanese Honorific Conversion](https://mlsanity.com/paper/2306.03055/) (2023)

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

- [When a Name Is Not a Name: A Benchmark Dataset and Distilled Reasoning for Culturally Entangled Bangla Homographs in Low-Resource LLMs](https://mlsanity.com/paper/2607.17828/) (2026)

## Abstract (arXiv)

Recent advances in Multilingual Large Language Models (MLLMs) have significantly enhanced cross-lingual conversational capabilities, yet modeling culturally nuanced and context-dependent communication remains a critical bottleneck. Specifically, existing state-of-the-art models exhibit a severe pragmatic gap when handling structural variations, regional idioms, and honorific consistencies in low-resource contexts like Bangla. To address this limitation, we introduce a novel, culturally aligned instruction-tuning dataset for BangLa Application and DialoguE generation - BLADE and benchmarking framework comprising 4,196 meticulously curated interaction pairs. We leverage this resource to systematically fine-tune and evaluate leading open-weight architectures, including DeepSeek-8B and LLaMA-3.2-3B, utilizing parameter-efficient fine-tuning via LoRA adapters in a 4-bit NormalFloat (NF4) quantization framework. Our empirical evaluations demonstrate that models fine-tuned on our dataset yield substantial improvements in structural fidelity and honorific alignment, providing a rigorous benchmark for bridging pragmatic disparities in low-resource multilingual text generation. Code and dataset: https://github.com/ashuvo25/BanglaApplicationLLM/tree/main