Results for "true negative rate"

Prevents attention to future tokens during training/inference.

Measures joint variability between variables.

Two-network setup where generator fools a discriminator.

Normalized covariance.

Measure of vector magnitude; used in regularization and optimization.

Flat high-dimensional regions slowing training.

Choosing step size along gradient direction.

Using output to adjust future inputs.

Regulating access to large-scale compute.

Configuration choices not learned directly (or not typically learned) that govern training or architecture.

A gradient method using random minibatches for efficient training on large datasets.

Uses an exponential moving average of gradients to speed convergence and reduce oscillation.

How many requests or tokens can be processed per unit time; affects scalability and cost.

A model is PAC-learnable if it can, with high probability, learn an approximately correct hypothesis from finite samples.

Converting audio speech into text, often using encoder-decoder or transducer architectures.

Direction of steepest ascent of a function.

Classical controller balancing responsiveness and stability.

Mathematical representation of friction forces.

A scalar measure optimized during training, typically expected loss over data, sometimes with regularization terms.

A function measuring prediction error (and sometimes calibration), guiding gradient-based optimization.

Minimizing average loss on training data; can overfit when data is limited or biased.

Separating data into training (fit), validation (tune), and test (final estimate) to avoid leakage and optimism bias.

Fraction of correct predictions; can be misleading on imbalanced datasets.

Variability introduced by minibatch sampling during SGD.

A hidden variable influences both cause and effect, biasing naive estimates of causal impact.

Diffusion model trained to remove noise step by step.

Estimating parameters by maximizing likelihood of observed data.

Generator produces limited variety of outputs.

Generative model that learns to reverse a gradual noise process.

Pixel-wise classification of image regions.