Protein Language Model Zero-Shot Fitness Predictions are Improved by Inference-only Dropout

Documenti analoghi

• Transformers as Unrolled Inference in Probabilistic Laplacian Eigenmaps: An Interpretation and Potential Improvements
  di: Ravuri, Aditya, et al.
  Pubblicazione: (2025)
• Towards One Model for Classical Dimensionality Reduction: A Probabilistic Perspective on UMAP and t-SNE
  di: Ravuri, Aditya, et al.
  Pubblicazione: (2024)
• Scalable Amortized GPLVMs for Single Cell Transcriptomics Data
  di: Zhao, Sarah, et al.
  Pubblicazione: (2024)
• No Generation without Representation: Efficient Causal Protein Language Models Enable Zero-Shot Fitness Estimation
  di: Eris, Furkan
  Pubblicazione: (2026)
• Predicting Language Models' Success at Zero-Shot Probabilistic Prediction
  di: Ren, Kevin, et al.
  Pubblicazione: (2025)