Arize AI, an ML observability platform, has launched Phoenix, an open-source library to monitor large language models (LLMs) for hallucinations.

Phoenix monitors false or misleading results in applications built on top of LLMs such as Google’s Bard and OpenAI’s GPT-4. It can be used to identify clusters of data problems and generate visuals for users to analyze, troubleshoot, or fine-tune the model to better results.

The product is designed as a standalone offering and runs locally on a notebook server.

Analyst QuickTake: Ever since the introduction of ChatGPT , a chatbot built on OpenAI's GPT model, there has been a growing interest in LLMs. These models have found widespread applications in areas such as generative and conversational AI , language translation and summarization, and synthetic data generation.