Results for "task-specific"

Recovering training data from gradients.

GNN framework where nodes iteratively exchange and aggregate messages from neighbors.

Simplified Boltzmann Machine with bipartite structure.

Diffusion model trained to remove noise step by step.

Models effects of interventions (do(X=x)).

What would have happened under different conditions.

Agent calls external tools dynamically.

Simple agent responding directly to inputs.

Scaling law optimizing compute vs data.

Models whose weights are publicly available.

Set of vectors closed under addition and scalar multiplication.

Measure of vector magnitude; used in regularization and optimization.

Probability of data given parameters.

Eliminating variables by integrating over them.

Visualization of optimization landscape.

Ensuring AI systems pursue intended human goals.

Explicit output constraints (format, tone).

Small prompt changes cause large output changes.

Requirement to inform users about AI use.

Ability to inspect and verify AI decisions.

Review process before deployment.

Assigning AI costs to business units.

Study of motion without considering forces.

Computing joint angles for desired end-effector pose.

Optimizing continuous action sequences.

Learning action mapping directly from demonstrations.

Interpreting human gestures.

AI systems assisting clinicians with diagnosis or treatment decisions.

Predicting disease progression or survival.

Software regulated as a medical device.