Recent research published on arXiv explores the use of trainable prompts within pretrained models to advance incremental learning, focusing specifically on incremental object detection. This approach leverages parameterized prompts to adapt models to new classes without retraining from scratch. However, current methods exhibit limitations, particularly in handling the co-occurrence of multiple classes within detection images. This shortcoming suggests that while trainable prompts show promise, their application in complex detection scenarios remains incomplete. The findings highlight a gap in existing techniques that future work may need to address to improve performance in real-world object detection tasks. These insights contribute to ongoing discussions in the computer vision community about optimizing incremental learning strategies. The research underscores the importance of refining prompt-based methods to better manage multi-class interactions in detection environments.