Results for "quadratic cost"

Optimal control for linear systems with quadratic cost.

Cost to run models in production.

Techniques to handle longer documents without quadratic cost.

Cost of model training.

Optimizing continuous action sequences.

Assigning AI costs to business units.

Optimal pathfinding algorithm.

Finding control policies minimizing cumulative cost.

Attention mechanisms that reduce quadratic complexity.

Restricting updates to safe regions.

Observing model inputs/outputs, latency, cost, and quality over time to catch regressions and drift.

Limiting inference usage.

Methods for breaking goals into steps; can be classical (A*, STRIPS) or LLM-driven with tool calls.

Optimizes future actions using a model of dynamics.

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

Of predicted positives, the fraction that are truly positive; sensitive to false positives.

Of true positives, the fraction correctly identified; sensitive to false negatives.

Of true negatives, the fraction correctly identified.

Selecting the most informative samples to label (e.g., uncertainty sampling) to reduce labeling cost.

When some classes are rare, requiring reweighting, resampling, or specialized metrics.

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

Optimization using curvature information; often expensive at scale.

Routes inputs to subsets of parameters for scalable capacity.

Increasing model capacity via compute.

Scaling law optimizing compute vs data.

Competitive advantage from proprietary models/data.

Visualization of optimization landscape.

Finding routes from start to goal.