Entity Tracking

Entity tracking, the task of monitoring how entities and their attributes change throughout a text or dialogue, is crucial for natural language understanding and numerous downstream applications. Current research focuses on improving the ability of large language models (LLMs) to perform this task, exploring techniques like pre-training on structured data (e.g., code), fine-tuning on diverse tasks, and incorporating external knowledge graphs to enhance reasoning capabilities. These advancements aim to create more robust and accurate models for applications such as question answering, summarization, and procedural text comprehension, ultimately improving the ability of machines to understand dynamic information within natural language.