L (29 terms)
Lagrangian
Converts constrained problem to unconstrained form.
Intermediate
Language Model
A model that assigns probabilities to sequences of tokens; often trained by next-token prediction.
Intermediate
Large Language Model
A high-capacity language model trained on massive corpora, exhibiting broad generalization and emergent behaviors.
Intermediate
Latency
Time from request to response; critical for real-time inference and UX.
Intermediate
Latency SLA
Guaranteed response times.
Intermediate
Latent Diffusion
Diffusion performed in latent space for efficiency.
Advanced
Latent Dynamics
Modeling environment evolution in latent space.
Frontier
Latent Space
The internal space where learned representations live; operations here often correlate with semantics or generative f...
Intermediate
Law of Large Numbers
Sample mean converges to expected value.
Advanced
Learning Rate
Controls the size of parameter updates; too high diverges, too low trains slowly or gets stuck.
Intermediate
Learning Rate Schedule
Adjusting learning rate over training to improve convergence.
Intermediate
Legal AI
AI supporting legal research, drafting, and analysis.
Intermediate
Legal Hold
Requirement to preserve relevant data.
Intermediate
Lifelong Learning
Learning without catastrophic forgetting.
Advanced
Likelihood Function
Probability of data given parameters.
Advanced
LIME
Local surrogate explanation method approximating model behavior near a specific input.
Intermediate
Line Search
Choosing step size along gradient direction.
Intermediate
Linear Algebra
Mathematical foundation for ML involving vector spaces, matrices, and linear transformations.
Advanced
Linear Quadratic Regulator
Optimal control for linear systems with quadratic cost.
Intermediate
Liquidity Provision
Supplying buy/sell orders.
Advanced
Local Minimum
Minimum relative to nearby points.
Intermediate
Localization
Estimating robot position within a map.
Advanced
Log Loss
Penalizes confident wrong predictions heavily; standard for classification and language modeling.
Intermediate
Logits
Raw model outputs before converting to probabilities; manipulated during decoding and calibration.
Intermediate
LoRA
PEFT method injecting trainable low-rank matrices into layers, enabling efficient fine-tuning.
Intermediate
Loss Function
A function measuring prediction error (and sometimes calibration), guiding gradient-based optimization.
Intermediate
Loss Landscape
The shape of the loss function over parameter space.
Intermediate
LSTM
An RNN variant using gates to mitigate vanishing gradients and capture longer context.
Intermediate
Lyapunov Stability
Stability proven via monotonic decrease of Lyapunov function.
Intermediate