🏷️:Cross-Lingual Auto Evaluation for Assessing Multilingual LLMs 🔗:https://t.co/OYTGde824b https://t.co/aq6T0b9UcC
🏷️:CBT-Bench: Evaluating Large Language Models on Assisting Cognitive Behavior Therapy 🔗:https://t.co/t0yMHNz3C4 https://t.co/p3axX4Ar9k
🏷️:Zero-shot Model-based Reinforcement Learning using Large Language Models 🔗:https://t.co/nGZX2LoYUw https://t.co/UOW1bEu8xf
Researchers from Carnegie Mellon University have introduced Pangea-7B, a fully open multilingual multimodal large language model (LLM) capable of understanding and responding in 39 languages. The model has been trained on a diverse dataset consisting of 6 million multilingual multimodal instruction tuning data, enabling it to outperform existing open-source multimodal LLMs, including notable models such as Llama-3.2-Vision, Molmo, and LlavaOneVision. Pangea-7B aims to bridge linguistic and cultural gaps in visual understanding, marking a significant advancement in inclusive artificial intelligence and language diversity.
