Prompting is the primary interface through which users interact with large language models. Although individual prompting strategies have been extensively studied, it remains unclear whether combining multiple strategies can yield additive benefits. We present a systematic empirical study of prompt pattern composition across multiple LLMs and composition mechanisms, evaluating 18 prompt pattern categories over 7 benchmarks spanning 6 task types. Among the three models for which we exhaustively evaluate all 306 ordered pairs of functional prompt patterns, only 6.86%, 10.46%, and 4.25% of compositions outperform their stronger constituent prompt, respectively. We further find that prompt accumulation provides only limited early-stage gains, while local additive predictions increasingly overestimate actual performance as compositions grow longer. Moreover, selection-based mechanisms generally provide a more favorable performance profile than direct prompt accumulation. Overall, our results reveal a scoped less-is-more tendency in the prompt-composition settings studied here: effective use of multiple prompt strategies depends more on selecting suitable strategies than on simply accumulating additional instructions.

Medical large language models hold promise for reducing healthcare disparities, yet Hindi remains severely underrepresented. While medical LLMs excel in high-resource languages, their performance degrades sharply in Hindi, particularly on Indian systems of medicine. We argue that robust cross-lingual medical transfer requires Hindi reasoning. To this end, we introduce HiMed, a Hindi reasoning medical corpus and benchmark suite covering both Western and Indian medicine. We further propose HiMed-8B, a Hindi-form medical reasoning LLM, through the design of decaying scaffolding reward. Extensive experiments demonstrate improvement in Hindi medical reasoning performance and a reduction in the English–Hindi accuracy gap. Ablation studies validate the contribution of each training stage and reward component. All code, data, and weights are available.

Controllable text-to-speech (TTS) aims to generate speech from text with precise control over prosodic and speaker-related attributes such as pitch, age, and accent. Existing methods mainly rely on natural language prompts or reference audio cloning. However, prompt-based approaches often suffer from the cross-modal semantic gap between textual descriptions and target speech attributes, resulting in imprecise and coarse-grained control, while cloning-based methods are constrained by reference audio and offer limited flexibility. To address these limitations, we propose TTS-Hub, a controllable TTS framework that leverages modular Low-Rank Adaptation (LoRA) components and their arithmetic composition for fine-grained and flexible speech control. Specifically, we first build a comprehensive Data Hub covering 6 high-level categories and 32 fine-grained speech attributes, and use it to construct a LoRA Hub in which each LoRA is specialized for a specific attribute. During inference, TTS-Hub selects and composes the required LoRA modules through simple arithmetic operations, enabling flexible and extensible attribute control without retraining the backbone model. Experiments on two mainstream TTS frameworks demonstrate that TTS-Hub consistently outperforms strong baselines and enables precise and flexible control.