sancovp/dark-factory-live - 5페이지

0-4-2-understand-context-length

[0.4.2] Total number of tokens (input + output) an LLM supports; the hard ceiling on how much text can be loaded into

원문 언어: 영어

0-4-1-understand-context-window

[0.4.1] The finite token sequence an LLM can process in a single forward pass; defines the maximum input + output capa

원문 언어: 영어

0-4-3-understand-long-term-memory

[0.4.3] Information persisted across sessions, tasks, or deployments; requires external storage and retrieval mechanis

원문 언어: 영어

0-4-6-understand-vector-database

[0.4.6] Storage system indexing embeddings for similarity search; backbone of RAG. Examples: Pinecone, Weaviate, Chrom

원문 언어: 영어

0-4-4-understand-working-memory

[0.4.4] The transient token budget allocated for active reasoning within a single model call; scratch space for chain_

원문 언어: 영어

0-1-1-understand-is-a

[0.1.1] A taxonomic relationship linking a kg_instance to its kg_class, establishing membership and enabling inheritan

원문 언어: 영어

0-1-understand-core

[0.1] the core subdomain of llm memory and context engineering (1 concepts)

원문 언어: 영어

0-2-1-understand-token-boundary-marker

[0.2.1] The positional delimiter marking the transition from input context to output generation; tokens before this ma

원문 언어: 영어

0-2-2-understand-input-token-budget

[0.2.2] The portion of context_length reserved and available for prompt tokens, retrieved context, and system instruct

원문 언어: 영어

0-2-3-understand-output-token-limit

[0.2.3] The portion of context_length reserved for generated tokens; equals context_length minus input_token_budget mi

원문 언어: 영어

0-2-4-understand-overflow-threshold

[0.2.4] The position or token count at which the model behavior shifts from normal processing to truncation or error;

원문 언어: 영어

0-2-5-understand-maximum-position

[0.2.5] The highest valid integer position index representable by the position_embedding scheme within context_length;

원문 언어: 영어

0-2-6-understand-context-capacity-ratio

[0.2.6] The ratio of tokens actively processed to total available context_length; quantifies context_utilization as us

원문 언어: 영어

0-2-understand-deep-context-window

[0.2] the deep_context_window subdomain of llm memory and context engineering (70 concepts)

원문 언어: 영어

0-3-1-understand-em-event

[0.3.1] A discrete unit of experience stored in episodic memory — a single interaction, task completion, or notable ou

원문 언어: 영어

0-3-2-understand-kg-entity

[0.3.2] A node in the knowledge_graph representing a distinct thing — concrete object, person, event, or concept — ide

원문 언어: 영어

0-3-3-understand-em-salience-weight

[0.3.3] Numeric value indicating how memorable or significant an event is; influences consolidation priority and retri

원문 언어: 영어

0-3-4-understand-kg-predicate

[0.3.4] The label or type of a kg_relationship; defines the semantics of how two kg_entities relate (e.g., is_a, part_

원문 언어: 영어

0-3-5-understand-kg-ontology

[0.3.5] The schema layer of a knowledge_graph defining allowed kg_entity types (classes), allowed kg_predicate labels,

원문 언어: 영어

0-3-6-understand-kg-class

[0.3.6] A kg_entity type within the kg_ontology; groups instances that share structural or semantic properties

원문 언어: 영어

0-3-understand-deep-long-term-memory

[0.3] the deep_long_term_memory subdomain of llm memory and context engineering (24 concepts)

원문 언어: 영어

0-4-1-understand-context-window

[0.4.1] The finite token sequence an LLM can process in a single forward pass; defines the maximum input + output capa

원문 언어: 영어

0-4-2-understand-context-length

[0.4.2] Total number of tokens (input + output) an LLM supports; the hard ceiling on how much text can be loaded into

원문 언어: 영어

0-4-3-understand-long-term-memory

[0.4.3] Information persisted across sessions, tasks, or deployments; requires external storage and retrieval mechanis

원문 언어: 영어

0-4-4-understand-working-memory

[0.4.4] The transient token budget allocated for active reasoning within a single model call; scratch space for chain_

원문 언어: 영어

0-4-5-understand-attention-mechanism

[0.4.5] The core operation allowing each token to attend to all others; scaled dot-product attention is the standard.

원문 언어: 영어

0-4-6-understand-vector-database

[0.4.6] Storage system indexing embeddings for similarity search; backbone of RAG. Examples: Pinecone, Weaviate, Chrom

원문 언어: 영어

0-4-understand-llm-memory-and-context-engineering

[0.4] the llm_memory_and_context_engineering subdomain of llm memory and context engineering (343 concepts)

원문 언어: 영어

0-understand-llm-memory-and-context-engineering

[0] the llm memory and context engineering library — every skill a certified region of the KB; call number = home : facets

원문 언어: 영어

using-llm-memory-and-context-engineering

Use the llm memory and context engineering neurosymbolic module: RAG library, agent brain, growable KB — proof-checked

원문 언어: 영어

0-understand-vector-databases-and-embeddings

[0] the vector databases and embeddings library — every skill a certified region of the KB; call number = home : facets

원문 언어: 영어

0-1-understand-core

[0.1] the core subdomain of vector databases and embeddings (30 concepts)

원문 언어: 영어

0-1-5-understand-authenticates-to-pinecone-project

[0.1.5] The relationship where a scoped API key grants authenticated access to a specific Pinecone project and its res

원문 언어: 영어

0-1-6-understand-belongs-to-pinecone-index

[0.1.6] The containment relationship indicating a record resides within a specific Pinecone index for storage and retr

원문 언어: 영어

0-1-1-understand-extends-pinecone-index

[0.1.1] The inheritance relationship where pod-based and serverless deployment models specialize the base Pinecone ind

원문 언어: 영어

0-1-2-understand-is-component-of-pinecone-record

[0.1.2] The partitive relationship where a dense or sparse vector constitutes one of the mandatory or optional fields

원문 언어: 영어

0-1-3-understand-reads-from-pinecone-index

[0.1.3] The data access relationship where fetch or query operations retrieve stored records from a Pinecone index by

원문 언어: 영어

0-1-4-understand-writes-to-pinecone-index

[0.1.4] The data mutation relationship where upsert or delete operations modify the record set stored within a Pinecon

원문 언어: 영어

0-2-understand-deep-nearest-neighbor-sea

[0.2] the deep_nearest_neighbor_sea subdomain of vector databases and embeddings (47 concepts)

원문 언어: 영어

0-2-5-understand-asymmetric-quantization

[0.2.5] A quantization scheme using both scale_factor and zero_point to map an arbitrary [min, max] float range to the

원문 언어: 영어