In this paper we present a novel methodology that harnesses the power of prompt tuning applied directly to Gemma3-12B, a state-of-the-art generative large language model to enhance performance on complex natural language processing challenges. Instead of relying solely on extensive retraining, our approach leverages carefully crafted input prompts to steer the pre-trained Gemma-12B towards generating outputs with superior contextual accuracy and interpretability. Our experimental evaluation employed a composite LLM Score metric that quantifies both semantic coherence and relevance; under this framework, our system (Team Name: Sarang) achieved a score of 4.54, ranking 9th in the shared task. Furthermore, in the competitive task evaluation, our method demonstrated the potential of prompt tuning as a viable alternative to traditional fine-tuning approaches. This study not only demonstrates the practical benefits of integrating prompt engineering with large language models but also opens avenues for future research aimed at further optimizing model performance in domain-specific applications.

Question-answering (QA) systems play a pivotal role in natural language processing (NLP), powering applications such as search engines and virtual assistants by providing accurate responses to user queries. However, building effective QA systems for Dravidian languages, like Tamil, poses distinct challenges due to the scarcity of resources and the linguistic complexities inherent to these languages. This paper introduces a novel method to enhance QA accuracy by integrating answer-type features alongside traditional question and context inputs. We fine-tuned both mono- and multilingual pre-trained models on the Extended Chaii dataset, which comprises Tamil translations from the SQuAD dataset, as well as on the SQuAD-EAT-5000 dataset, consisting of English-language instances. Our experiments reveal that incorporating answer-type features significantly improves model performance compared to using only question and context inputs. Specifically, for the Extended Chaii dataset, the MuRIL model achieved the highest F1 score of 53.89, surpassing other pre-trained models, while RoBERTa outperformed BERT on the SQuAD-EAT-5000 dataset with a score of 82.07. This research advances QA systems for Dravidian languages and underscores the importance of integrating linguistic features for improved accuracy.

News classification allows analysts and researchers to study trends over time. Based on classification, news platforms can provide readers with related articles. Many digital news platforms and apps use classification to offer personalized content for their users. While there are numerous resources accessible for news classification in various Indian languages, there is still a lack of extensive benchmark dataset specifically for the Telugu language. Our paper presents and describes the Telugu20news group dataset, where news has been collected from various online Telugu news channels. We describe in detail the accumulation and annotation of the proposed news headlines dataset. In addition, we conducted extensive experiments on our proposed news headlines dataset in order to deliver solid baselines for future work.