7 papers · 1 filter
Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension
Raviraj Joshi, Utkarsh Vaidya, Sanjay Singh Chauhan +1
Vocabulary extension is an efficient way to adapt pretrained large language models (LLMs) to new languages, but the initialization of newly added token embeddings can strongly affe…
Benchmarking Hindi LLMs: A New Suite of Datasets and a Comparative Analysis
Anusha Kamath, Kanishk Singla, Rakesh Paul +4
Evaluating instruction-tuned Large Language Models (LLMs) in Hindi is challenging due to a lack of high-quality benchmarks, as direct translation of English datasets fails to captu…
CultureGuard: Towards Culturally-Aware Dataset and Guard Model for Multilingual Safety Applications
Raviraj Joshi, Rakesh Paul, Kanishk Singla +8
The increasing use of Large Language Models (LLMs) in agentic applications highlights the need for robust safety guard models. While content safety in English is well-studied, non-…
Aligning Large Language Models to Low-Resource Languages through LLM-Based Selective Translation: A Systematic Study
Rakesh Paul, Anusha Kamath, Kanishk Singla +4
Multilingual large language models (LLMs) often demonstrate a performance gap between English and non-English languages, particularly in low-resource settings. Aligning these model…
On Importance of Code-Mixed Embeddings for Hate Speech Identification
Shruti Jagdale, Omkar Khade, Gauri Takalikar +2
Code-mixing is the practice of using two or more languages in a single sentence, which often occurs in multilingual communities such as India where people commonly speak multiple l…
Challenges in Adapting Multilingual LLMs to Low-Resource Languages using LoRA PEFT Tuning
Omkar Khade, Shruti Jagdale, Abhishek Phaltankar +2
Large Language Models (LLMs) have demonstrated remarkable multilingual capabilities, yet challenges persist in adapting these models for low-resource languages. In this study, we i…