Domain: AI Economics & Strategy

Mechanics of price formation.

Formal framework for sequential decision-making under uncertainty.

Estimating parameters by maximizing likelihood of observed data.

Extending agents with long-term memory stores.

Routes inputs to subsets of parameters for scalable capacity.

Central catalog of deployed and experimental models.

Inferring sensitive features of training data.

Competitive advantage from proprietary models/data.

Coordinating models, tools, and logic.

Risk of incorrect financial models.

Framework for identifying, measuring, and mitigating model risks.

Embedding signals to prove model ownership.

Multiple agents interacting cooperatively or competitively.

Allows model to attend to information from different subspaces simultaneously.

Quantifies shared information between random variables.

Optimization with multiple local minima/saddle points; typical in neural networks.

Learning from data generated by a different policy.

Learning only from current policy’s data.

Models whose weights are publicly available.

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

Using same parameters across different parts of a model.

Separates planning from execution in agent architectures.

Strategy mapping states to actions.

Optimizing policies directly via gradient ascent on expected reward.

Extracting system prompts or hidden instructions.

Expected return of taking action in a state.

Measures a model’s ability to fit random noise; used to bound generalization error.

Allows gradients to bypass layers, enabling very deep networks.

Quantifying financial risk.

Encodes positional information via rotation in embedding space.