Mi:dm K 2.5 Pro

TL;DR

Mi:dm K 2.5 Pro is a large, reasoning-focused language model optimized for enterprise and Korean-language tasks, featuring advanced training techniques and safety measures to enhance complex problem-solving and cultural understanding.

Contribution

The paper introduces Mi:dm K 2.5 Pro, a 32B parameter LLM with novel training strategies like Depth Upscaling and Fusion Training, tailored for enterprise complexity and Korean language benchmarks.

Findings

Achieves state-of-the-art results on Korean benchmarks.

Demonstrates strong reasoning and problem-solving capabilities.

Ensures safety and responsible AI deployment.

Abstract

The evolving LLM landscape requires capabilities beyond simple text generation, prioritizing multi-step reasoning, long-context understanding, and agentic workflows. This shift challenges existing models in enterprise environments, especially in Korean-language and domain-specific scenarios where scaling is insufficient. We introduce Mi:dm K 2.5 Pro, a 32B parameter flagship LLM designed to address enterprise-grade complexity through reasoning-focused optimization. Our methodology builds a robust data foundation via a quality-centric curation pipeline utilizing abstract syntax tree (AST) analysis for code, gap-filling synthesis for mathematics, and an LLM-based quality evaluator. Pre-training scales the model via layer-predictor-based Depth Upscaling (DuS) and a progressive strategy supporting a 128K token context window. Post-training introduces a specialized multi-stage pipeline, including Reasoning SFT, model merging, and asynchronous reinforcement learning (RL), to develop complex problem-solving skills. "Fusion Training" then rebalances these capabilities with conversational fluency, consistent response styling, and reliable tool-use. The evaluations show that Mi:dm K 2.5 Pro achieves competitive performance against leading global and domestic models. In addition, it sets state-of-the-art results on Korean-specific benchmarks, showcasing deep linguistic and cultural understanding. Finally, Responsible AI evaluations validate safety against attacks, ensuring a secure profile for deployment with a balance of harmlessness and responsiveness.