ARCHIE Goes to NeurIPS2025!
2025-10-02
Our paper “AudSemThinker: Enhancing Audio-Language Models through Reasoning over Semantics of Sound” has been accepted at NeurIPS2025 in San Diego! 🎉

2025-10-02
Our paper “AudSemThinker: Enhancing Audio-Language Models through Reasoning over Semantics of Sound” has been accepted at NeurIPS2025 in San Diego! 🎉
The paper by Gijs Wingaard, Elia Formisano, Michele Esposito and Michel Dumontier introduces AudSemThinker, a model whose reasoning is structured around a framework of auditory semantics inspired by human cognition and AudSem, a novel dataset specifically curated for semantic descriptor reasoning in audio-language models.
You can read the preprint here