Sound has always held a profound significance in the realm of spiritual practices, with music and chants often used as tools for connecting with the divine. This age-old tradition can be seen in the mystical world of Sufism, where music and sound are not just forms of entertainment, but pathways to spiritual enlightenment.
At the Musée d’Art et de Culture Soufis MTO, an upcoming exhibition delves into the intricate relationship between Sufism and sound. This unique exploration showcases how music, poetry, and rituals are used in Sufi practices to induce states of ecstasy and transcendence.
From the mesmerizing melodies of the qawwali music to the rhythmic whirling of the dervishes, the exhibition highlights the diverse ways in which sound is utilized in Sufism. Visitors will have the opportunity to immerse themselves in the rich tapestry of Sufi traditions, gaining a deeper understanding of the spiritual significance of sound.
As we navigate the complexities of the modern world, the exhibition serves as a reminder of the timeless wisdom found in ancient practices. By exploring the intersection of Sufism and sound, we are invited to reflect on the universal power of music and its ability to transcend cultural boundaries.
New features in the Responses API: Remote MCP, image gen, Code Interpreter, and more. Powering faster, smarter agents with GPT-4o & o-series models, plus new features for reliability and efficiency.
arXiv:2505.06149v2 Announce Type: replace-cross Abstract: Despite growing interest in automated hate speech detection, most existing approaches overlook the linguistic diversity of online content. Multilingual instruction-tuned large language models such as LLaMA, Aya, Qwen, and BloomZ offer promising capabilities across languages, but their effectiveness in identifying hate speech through zero-shot and few-shot prompting remains underexplored. This work evaluates LLM prompting-based detection across eight non-English languages, utilizing several prompting techniques and comparing them to fine-tuned encoder models. We show that while zero-shot and few-shot prompting lag behind fine-tuned encoder models on most of the real-world evaluation sets, they achieve better generalization on functional tests for hate speech detection. Our study also reveals that prompt design plays a critical role, with each language often requiring customized prompting techniques to maximize performance.
Building a family of AI products for everyone.