Results for "full pass through data"

Tradeoffs between many layers vs many neurons per layer.

Assigning labels per pixel (semantic) or per instance (instance segmentation) to map object boundaries.

Strategy mapping states to actions.

Expected return of taking action in a state.

Combines value estimation (critic) with policy learning (actor).

Coordination arising without explicit programming.

Required human review for high-risk decisions.

Extracting system prompts or hidden instructions.

Embedding signals to prove model ownership.

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

Joint vision-language model aligning images and text.

Formal model linking causal mechanisms and variables.

Probability of treatment assignment given covariates.

Variable enabling causal inference despite confounding.

Agent reasoning about future outcomes.

Declining differentiation among models.

Organizational uptake of AI technologies.

Matrix of curvature information.

Variable whose values depend on chance.

Ensuring AI systems pursue intended human goals.

Model exploits poorly specified objectives.

Tendency for agents to pursue resources regardless of final goal.

Model optimizes objectives misaligned with human values.

Ensuring learned behavior matches intended objective.

Learned subsystem that optimizes its own objective.

Assigning a role or identity to the model.

Explicit output constraints (format, tone).

Breaking tasks into sub-steps.

Differences between training and inference conditions.

Governance of model changes.