V (15 terms)

Value at Risk Maximum expected loss under normal conditions. Intermediate Value Function Expected cumulative reward from a state or state-action pair. Intermediate Value Learning Inferring and aligning with human preferences. Intermediate Value Misalignment Model optimizes objectives misaligned with human values. Advanced Vanishing Gradient Gradients shrink through layers, slowing learning in early layers; mitigated by ReLU, residuals, normalization. Intermediate Variance Measure of spread around the mean. Advanced Variance Term Error due to sensitivity to fluctuations in the training dataset. Intermediate Variational Autoencoder Autoencoder using probabilistic latent variables and KL regularization. Advanced VC Dimension A measure of a model class’s expressive capacity based on its ability to shatter datasets. Intermediate Vector Database A datastore optimized for similarity search over embeddings, enabling semantic retrieval at scale. Intermediate Vector Space Set of vectors closed under addition and scalar multiplication. Advanced Vickrey Auction Truthful bidding is optimal strategy. Advanced Vision Transformer Transformer applied to image patches. Intermediate Vocabulary The set of tokens a model can represent; impacts efficiency, multilinguality, and handling of rare strings. Intermediate Voice Conversion Changing speaker characteristics while preserving content. Intermediate