发表机构
LG AI Research（LG AI研究院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该报告介绍LG AI Research开发的K-EXAONE 2.0，这是一款7500亿参数的MoE多语言基础模型，经升级前代模型而来，支持25.6万token上下文，在多类评估中表现优异，以Apache 2.0许可发布，助力AI生态发展。
AI 中文摘要
本技术报告介绍了由LG AI Research开发的开放权重多语言基础模型K-EXAONE 2.0，是我们打造全球前沿级基础模型努力中的一步。我们未从头训练，而是对K-EXAONE进行升级并扩展其架构，得到了一个混合专家（Mixture-of-Experts，MoE）模型，总参数达7500亿，每个token激活约370亿，容量是前代的三倍多。K-EXAONE 2.0支持最长25.6万个token的上下文长度，多语言覆盖从6种扩展到10种。其训练流程结合了持续预训练、聚焦难度的中期训练和后训练，以强化推理、智能体编码、多语言能力以及基于韩国社会文化语境的安全性。在9个反映实际使用场景的评估类别中，K-EXAONE 2.0较K-EXAONE有所提升，且与其他开放权重模型保持竞争力，在智能体编码和长上下文理解方面提升最大，在长上下文检索和安全性方面优势最明显。K-EXAONE 2.0以Apache 2.0许可发布，使更广泛的AI生态系统能够对其进行评估、部署、适配和基于它构建，同时标志着我们向全球前沿发起挑战的开端而非终点。
英文摘要
This technical report presents K-EXAONE 2.0, an open-weight multilingual foundation model developed by LG AI Research as a step in our effort toward global frontier-scale foundation models. Rather than training from scratch, we upcycle K-EXAONE and expand its architecture, yielding a Mixture-of-Experts (MoE) model with 750B total parameters and approximately 37B activated per token---more than three times the capacity of its predecessor. K-EXAONE 2.0 supports context lengths of up to 256K tokens and expands multilingual coverage from six to ten languages. Its training pipeline combines continual pre-training, difficulty-focused mid-training, and post-training to strengthen reasoning, agentic coding, multilingual capability, and safety grounded in Korean sociocultural contexts. Across nine evaluation categories selected to reflect the conditions of practical use, K-EXAONE 2.0 improves over K-EXAONE and remains competitive with open-weight models, showing its largest gains in agentic coding and long-context understanding and its clearest strengths in long-context retrieval and safety. Released under the Apache 2.0 license, K-EXAONE 2.0 enables the wider AI ecosystem to evaluate, deploy, adapt, and build upon it, while marking the beginning---rather than the endpoint---of our challenge toward the global frontier.
