B. Bal

eGrantha.ai at PAN 2026: Contrastive Representation Learning for Robust Generative AI Authorship Detection Notebook for the PAN Lab at CLEF 2026

Nepal Script Text Recognition from Ancient Artifacts: Challenges and Opportunities

Nepali Legal Expertise through Generative and Extractive Pre-trained Transformers (NepLEGiT)

NepLEGiT (Nepali Legal Expertise through Generative and Extractive Pre-trained Transformers), a specialized small language model (SLM) designed to democratize legal knowledge and enhance legal-service delivery in Nepal.

NepKANUN: A RAG-Based Nepali Legal Assistant

An AI-powered legal assistant that is tailored for Nepali legal texts and is built on a fine-tuned large language model is introduced, showing how merging generation and retrieval can effectively democratize access to legal knowledge in Nepal by focusing on customized legal data and incorporating RAG.

Nepali Passport Question Answering: A Low-Resource Dataset for Public Service Applications

This study prepares a pair-structured Nepali Question-Answer dataset, focusing on Frequently Asked Questions for passport-related services, and implements a hybrid retrieval approach, integrating fine-tuned models with BM25, and evaluates the performance of the hybrid retrieval.

Towards Building a Standard Benchmark for Low-Resource Nepali Information Retrieval

It is argued that Nepali provides a compelling testbed for studying morphology-aware retrieval, code-mixed query processing, and multilingual embedding transfer.