Deployed multimodal systems can fail in ways that evaluators did not anticipate. In order to find these failures before deployment, we introduce MultiMon, a system that automatically identifies systematic failures -- generalizable, natural-language descriptions of patterns of model failures. To uncover systematic failures, MultiMon scrapes a corpus for examples of erroneous agreement: inputs that produce the same output, but should not. It then prompts a language model (e.g., GPT-4) to find systematic patterns of failure and describe them in natural language. We use MultiMon to find 14 systematic failures (e.g., "ignores quantifiers") of the CLIP text-encoder, each comprising hundreds of distinct inputs (e.g., "a shelf with a few/many books"). Because CLIP is the backbone for most state-of-the-art multimodal systems, these inputs produce failures in Midjourney 5.1, DALL-E, VideoFusion, and others. MultiMon can also steer towards failures relevant to specific use cases, such as self-driving cars. We see MultiMon as a step towards evaluation that autonomously explores the long tail of potential system failures. Code for MULTIMON is available at https://github.com/tsb0601/MultiMon.

翻译：已部署的多模态系统可能以评估者未曾预料的方式失效。为了在部署前发现这些失效，我们引入了MultiMon——一种能自动识别系统性失效的系统，即模型失效模式的可泛化自然语言描述。为揭示系统性失效，MultiMon通过语料库搜集错误同意的示例：产生相同输出但本应不同的输入。随后，它引导语言模型（如GPT-4）寻找失效的系统性模式，并以自然语言描述。我们利用MultiMon发现了CLIP文本编码器的14种系统性失效（例如“忽略量词”），每种失效包含数百个不同的输入（如“一个放有少量/多本书的书架”）。由于CLIP是大多数最先进多模态系统的基础模型，这些输入导致Midjourney 5.1、DALL-E、VideoFusion等系统产生失效。MultiMon还能针对特定用例相关的失效进行定向引导，例如自动驾驶汽车。我们将MultiMon视作迈向自主探索潜在系统失效长尾分布的评估步骤。MultiMon的代码已开源，地址为https://github.com/tsb0601/MultiMon。