Results for "language"

A model that assigns probabilities to sequences of tokens; often trained by next-token prediction.

Controlling robots via language.

AI subfield dealing with understanding and generating human language, including syntax, semantics, and pragmatics.

A high-capacity language model trained on massive corpora, exhibiting broad generalization and emergent behaviors.

The set of tokens a model can represent; impacts efficiency, multilinguality, and handling of rare strings.

Predicts masked tokens in a sequence, enabling bidirectional context; often used for embeddings rather than generation.

The text (and possibly other modalities) given to an LLM to condition its output behavior.

Crafting prompts to elicit desired behavior, often using role, structure, constraints, and examples.

Models that process or generate multiple modalities, enabling vision-language tasks, speech, video understanding, etc.

Converting audio speech into text, often using encoder-decoder or transducer architectures.

Penalizes confident wrong predictions heavily; standard for classification and language modeling.

Maximum number of tokens the model can attend to in one forward pass; constrains long-document reasoning.

A branch of ML using multi-layer neural networks to learn hierarchical representations, often excelling in vision, speech, and language.

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

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

Fine-tuning on (prompt, response) pairs to align a model with instruction-following behaviors.

Automated detection/prevention of disallowed outputs (toxicity, self-harm, illegal instruction, etc.).

Exponential of average negative log-likelihood; lower means better predictive fit, not necessarily better utility.

Models trained to decide when to call tools.

Joint vision-language model aligning images and text.

Aligns transcripts with audio timestamps.

Temporal and pitch characteristics of speech.

Task instruction without examples.

Assigning a role or identity to the model.

AI supporting legal research, drafting, and analysis.

Updating a pretrained model’s weights on task-specific data to improve performance or adapt style/behavior.

The field of building systems that perform tasks associated with human intelligence—perception, reasoning, language, planning, and decision-making—via algori...

Using markers to isolate context segments.

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

Attention where queries/keys/values come from the same sequence, enabling token-to-token interactions.