Shiyun Xiong, Dongming Wu, Peiwen Sun, Bokang Yang, Jiaxu Wang, Xiaolou Sun, Mingxing Wen, Xiangyu Yue · 2026 · arXiv
DOI: 10.48550/arXiv.2610.26751 · open-access PDF
Os sistemas incorporados recentes baseiam-se cada vez mais em modelos multimodais de grande linguagem (MLLMs) por suas capacidades pré-treinadas de compreensão e raciocínio, mas sua percepção do mundo físico permanece em grande parte limitada a modalidades básicas, como visão e áudio. No entanto, ampliar os MLLMs para modalidades sensoriais mais diversas do ambiente físico é um desafio: suas representações são muitas vezes mal conectadas com o MLLM e variam substancialmente em estrutura e estatísticas, enquanto sua interpretação ainda requer compreensão e raciocínio de alto nível compartilhados.
Machine translation (Meta M2M100); the English original follows.
Recent embodied systems increasingly build on multimodal large language models (MLLMs) for their pretrained understanding and reasoning capabilities, but their perception of the physical world remains largely limited to basic modalities such as vision and audio. However, broadening MLLMs to more diverse sensory modalities from the physical environment is challenging: their representations are often weakly connected to the MLLM and vary substantially in structure and statistics, while their interpretation still requires shared high-level understanding and reasoning. We therefore propose Unisens
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26751, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange