In our newest preprint, we discuss current explainable AI (XAI) methods. We divided the workflow of a generative decoder-only model into four information contexts for XAI: training dataset, input query, model components, and output sequence. See here: arxiv.org/abs/2506.19532
@aichemist.bsky.social

Are you curious about what protein language models learn? Check out our newest preprint! 🚀https://arxiv.org/abs/2506.19532
We reviewed explainable AI (XAI) techniques across all parts of the generative protein design workflow and discussed their applications, limitations, and untapped potential!