📄 中文摘要
深度学习模型的复杂性使其在决策过程中的透明度受到质疑。解释方法的出现旨在提高模型的可解释性，帮助用户理解模型的决策依据。不同的用户群体，如研究人员、开发者和最终用户，对模型解释的需求各不相同，价值观和关注点也有所不同。尽管已有多种解释方法被提出，但在实际应用中仍面临诸多挑战，包括解释的准确性、可用性和对用户的适应性等。因此，如何在保证模型性能的同时，提升其可解释性，成为当前深度学习领域的重要研究方向。
深度学习中的解释方法：用户、价值、关注点与挑战
📄 English Summary
Explanation Methods in Deep Learning: Users, Values, Concerns and Challenges
The complexity of deep learning models raises concerns about their transparency in decision-making processes. Explanation methods have emerged to enhance model interpretability, aiding users in understanding the rationale behind decisions. Different user groups, such as researchers, developers, and end-users, have varying needs regarding model explanations, along with differing values and concerns. Despite the introduction of various explanation methods, practical applications still face numerous challenges, including the accuracy, usability, and adaptability of explanations to users. Consequently, improving model interpretability while maintaining performance has become a significant research focus in the field of deep learning.