Behind SEA-LION: Pushing the Boundaries of Southeast Asian LLM capabilities

​Join us for an exclusive behind-the-scenes look at how the SEA-LION team builds and advances large language models for Southeast Asia. In this session, you’ll hear directly from the engineers who trained SEA-LION models, including the latest SEA-LION v3 and v4 built on  SOTA models - Gemma, Llama and Qwen models, and get an inside view of the full development pipeline—from data pipeline and large-scale pre-training to post-training, model alignment, and real-world applications.

​Attendees will also gain practical insights into the key technical decisions, trade-offs, and lessons learned in creating high-quality, regionally grounded LLMs to push the boundaries of performance and usability across Southeast Asian languages.