Results for "unequal knowledge"

Unequal performance across demographic groups.

Structured graph encoding facts as entity–relation–entity triples.

Agents communicate via shared state.

Loss of old knowledge when learning new tasks.

Belief before observing data.

Reusing knowledge from a source task/domain to improve learning on a target task/domain, typically via pretrained models.

Constraining outputs to retrieved or provided sources, often with citation, to improve factual reliability.

Task instruction without examples.

Learning without catastrophic forgetting.

AI proposing scientific hypotheses.

AI selecting next experiments.

Awareness and regulation of internal processes.

Training a smaller “student” model to mimic a larger “teacher,” often improving efficiency while retaining performance.

Letting an LLM call external functions/APIs to fetch data, compute, or take actions, improving reliability.

Designing input features to expose useful structure (e.g., ratios, lags, aggregations), often crucial outside deep learning.

Retrieval based on embedding similarity rather than keyword overlap, capturing paraphrases and related concepts.

Model-generated content that is fluent but unsupported by evidence or incorrect; mitigated by grounding and verification.

Framework for reasoning about cause-effect relationships beyond correlation, often using structural assumptions and experiments.

Mechanisms for retaining context across turns/sessions: scratchpads, vector memories, structured stores.

Quantifies shared information between random variables.

Built-in assumptions guiding learning efficiency and generalization.

Routes inputs to subsets of parameters for scalable capacity.

Learning from data generated by a different policy.

Extending agents with long-term memory stores.

Models trained to decide when to call tools.

Graphs containing multiple node or edge types with different semantics.

Simultaneous Localization and Mapping for robotics.

Interleaving reasoning and tool use.

Agent reasoning about future outcomes.

Prompt augmented with retrieved documents.