Recent advancements in Large Language Models (LLMs) have shown promising results in music understanding and generation tasks. However, existing works remain confined to Western tonal traditions, offering little insight into whether current LLMs can handle structurally distinct low-resource musical traditions. We present the first systematic evaluation of LLM competence in South Asian classical music, a tradition governed by raga, tala-based melodic constraints that impose fundamentally different structural principles from Western harmony-driven music. We ground our evaluation in Hindustani classical theory and Bengali classical forms, including Rabindra and Nazrul Sangeet – representative low-resource traditions within South Asian classical music. For music understanding evaluation, we introduce a 504-question-answer benchmark spanning raga grammar, cultural knowledge, and symbolic notation reasoning, evaluating 33 LLMs where frontier models such as Gemini 2.5 Pro achieve 85-90% accuracy, while most open-source models remain in the 23-40% range. For music generation, we design a five-level controlled prompting framework and find that even the strongest model produces stylistically faithful outputs only 40% of the time. These results reveal that structural validity and stylistic faithfulness in music generation are distinct objectives and highlight an open challenge for culturally grounded music modeling.

Sentence encoders are foundational to a wide range of NLP applications, yet they are predominantly evaluated through performance on downstream tasks. Such task-specific evaluations provide limited insight into whether the resulting sentence representations capture fundamental compositional relationships in a task-independent manner. To address this gap, we introduce a set-theoretic framework for evaluating the compositional properties of sentence embeddings. Drawing on classical set theory, we define six evaluation criteria based on three core set-like text operations: TextOverlap, TextDifference, and TextUnion. We use these criteria to systematically evaluate 19 sentence encoders, including 7 classical models and 12 Large Language Model (LLM)-based encoders. Our results reveal substantial differences in their compositional behavior: notably, SBERT exhibits the strongest overall alignment with the proposed set-theoretic criteria, outperforming even recent LLM-based encoders. We further examine whether these intrinsic compositional properties are associated with downstream performance across five NLP tasks. Among them, summarization exhibits the strongest correlation with the TextOverlap and TextDifference criteria, suggesting that our framework may also provide useful signals about downstream capability without relying exclusively on task-specific evaluation. Finally, we release a benchmark comprising ~192K samples to support systematic evaluation and future research on the compositional properties of sentence embeddings.

Sarcasm is a form of figurative language that serves as a humorous tool for mockery and ridicule. We present a novel architecture for sarcasm generation with emoji from a non-sarcastic input sentence in English. We divide the generation task into two sub tasks: one for generating textual sarcasm and another for collecting emojis associated with those sarcastic sentences. Two key elements of sarcasm are incorporated into the textual sarcasm generation task: valence reversal and semantic incongruity with context, where the context may involve shared commonsense or general knowledge between the speaker and their audience. The majority of existing sarcasm generation works have focused on this textual form. However, in the real world, when written texts fall short of effectively capturing the emotional cues of spoken and face-to-face communication, people often opt for emojis to accurately express their emotions. Due to the wide range of applications of emojis, incorporating appropriate emojis to generate textual sarcastic sentences helps advance sarcasm generation. We conclude our study by evaluating the generated sarcastic sentences using human judgement. All the codes and data used in this study has been made publicly available.