Command A: An Enterprise-Ready Large Language Model

In this report we describe the development of Command A, a powerful large language model purpose-built to excel at real-world enterprise use cases. Command A is an agent-optimised and multilingual-capable model, with support for 23 languages of global business, and a novel hybrid architecture balancing efficiency with top of the range performance. It offers best-in-class Retrieval Augmented Generation (RAG) capabilities with grounding and tool use to automate sophisticated business processes. These abilities are achieved through a decentralised training approach, including self-refinement algorithms and model merging techniques. We also include results for Command R7B which shares capability and architectural similarities to Command A. Weights for both models have been released for research purposes. This technical report details our original training pipeline and presents an extensive evaluation of our models across a suite of enterprise-relevant tasks and public benchmarks, demonstrating excellent performance and efficiency.

RICoTA: Red-teaming of In-the-wild Conversation with Test Attempts

User interactions with conversational agents (CAs) evolve in the era of heavily guardrailed large language models (LLMs). As users push beyond programmed boundaries to explore and build relationships with these systems, there is a growing concern regarding the potential for unauthorized access or manipulation, commonly referred to as "jailbreaking." Moreover, with CAs that possess highly human-like qualities, users show a tendency toward initiating intimate sexual interactions or attempting to tame their chatbots. To capture and reflect these in-the-wild interactions into chatbot designs, we propose RICoTA, a Korean red teaming dataset that consists of 609 prompts challenging LLMs with in-the-wild user-made dialogues capturing jailbreak attempts. We utilize user-chatbot conversations that were self-posted on a Korean Reddit-like community, containing specific testing and gaming intentions with a social chatbot. With these prompts, we aim to evaluate LLMs' ability to identify the type of conversation and users' testing purposes to derive chatbot design implications for mitigating jailbreaking risks. Our dataset will be made publicly available via GitHub.

DMOps: Data Management Operation and Recipes

Data-centric AI has shed light on the significance of data within the machine learning (ML) pipeline. Acknowledging its importance, various research and policies are suggested by academia, industry, and government departments. Although the capability of utilizing existing data is essential, the capability to build a dataset has become more important than ever. In consideration of this trend, we propose a "Data Management Operation and Recipes" that will guide the industry regardless of the task or domain. In other words, this paper presents the concept of DMOps derived from real-world experience. By offering a baseline for building data, we want to help the industry streamline its data operation optimally.