{"schemaVersion":1,"publisher":"Loaded Tokens","updated":"2026-11-07","terms":[{"slug":"agent","term":"Agent","definition":"A system where a model dynamically chooses steps or tools toward a goal, within the application’s constraints.","source":"https://www.anthropic.com/engineering/building-effective-agents","learn":"https://loadedtokens.com/learn/agents-versus-workflows/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#agent"},{"slug":"ai","term":"AI","definition":"Artificial intelligence: an umbrella term for systems performing tasks associated with intelligent behavior. It is broader than chatbots.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/what-a-language-model-does/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#ai"},{"slug":"batch-processing","term":"Batch processing","definition":"Submitting work under a provider’s asynchronous batch mode. Pricing and completion requirements differ from interactive calls.","source":"https://developers.openai.com/api/docs/pricing","learn":"https://loadedtokens.com/learn/evals-and-cost-per-success/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#batch-processing"},{"slug":"benchmark","term":"Benchmark","definition":"A defined task set and scoring setup used to compare performance. Results depend on conditions and may not match your workflow.","source":"https://platform.openai.com/docs/guides/evals","learn":"https://loadedtokens.com/learn/choose-a-model-by-the-job/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#benchmark"},{"slug":"checkpoint","term":"Checkpoint","definition":"A saved execution state that can support inspection or resuming a workflow. It is not a guarantee of safe behavior.","source":"https://docs.langchain.com/oss/python/langgraph/overview","learn":"https://loadedtokens.com/learn/agents-versus-workflows/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#checkpoint"},{"slug":"chunk","term":"Chunk","definition":"A passage or segment prepared for retrieval. Boundaries affect which context is available around a matched passage.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#chunk"},{"slug":"context-window","term":"Context window","definition":"The model’s finite working context for a request. Available input and output capacity depends on the model and product.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/tokens-context-and-memory/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#context-window"},{"slug":"cost-per-successful-task","term":"Cost per successful task","definition":"Total workflow spend divided by outcomes meeting the declared acceptance criteria, including failed attempts in the spend.","source":"https://platform.openai.com/docs/guides/evals","learn":"https://loadedtokens.com/learn/evals-and-cost-per-success/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#cost-per-successful-task"},{"slug":"data-governance","term":"Data governance","definition":"The policies and responsibilities controlling how data is accessed, used, retained and managed in a system.","source":"https://www.nist.gov/itl/ai-risk-management-framework","learn":"https://loadedtokens.com/learn/safe-workplace-ai/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#data-governance"},{"slug":"embedding","term":"Embedding","definition":"A numerical vector representing content so relatedness can be compared for tasks such as search.","source":"https://platform.openai.com/docs/guides/embeddings","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#embedding"},{"slug":"evaluation","term":"Evaluation","definition":"A repeatable assessment of outputs or system behavior against explicit criteria and representative examples.","source":"https://platform.openai.com/docs/guides/evals","learn":"https://loadedtokens.com/learn/evals-and-cost-per-success/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#evaluation"},{"slug":"evaluation-set","term":"Evaluation set","definition":"A collection of cases used to check behavior. Preserve versions and avoid treating a small set as universal proof.","source":"https://platform.openai.com/docs/guides/evals","learn":"https://loadedtokens.com/learn/evals-and-cost-per-success/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#evaluation-set"},{"slug":"few-shot-prompting","term":"Few-shot prompting","definition":"Supplying a small set of examples in the prompt to demonstrate the desired input/output behavior.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/prompting-with-a-brief/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#few-shot-prompting"},{"slug":"fine-tuning","term":"Fine-tuning","definition":"Further training a model on additional examples to alter its behavior. It is not the same as retrieving documents at request time.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#fine-tuning"},{"slug":"generative-ai","term":"Generative AI","definition":"Systems that generate outputs such as text or other media rather than only selecting a fixed label.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/what-a-language-model-does/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#generative-ai"},{"slug":"grounding","term":"Grounding","definition":"Connecting a response to supplied or retrieved evidence. The evidence still needs to support the actual claim.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#grounding"},{"slug":"guardrail","term":"Guardrail","definition":"A control intended to constrain inputs, outputs or actions. It reduces specific risks; it is not proof of universal safety.","source":"https://www.nist.gov/itl/ai-risk-management-framework","learn":"https://loadedtokens.com/learn/safe-workplace-ai/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#guardrail"},{"slug":"hallucination","term":"Hallucination","definition":"A plausible-looking output that is false or unsupported. Fluency alone is not evidence of correctness.","source":"https://www.nist.gov/itl/ai-risk-management-framework","learn":"https://loadedtokens.com/learn/what-a-language-model-does/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#hallucination"},{"slug":"human-in-the-loop","term":"Human in the loop","definition":"A workflow that includes a person at a defined review, approval or correction point.","source":"https://docs.langchain.com/oss/python/langgraph/overview","learn":"https://loadedtokens.com/learn/agents-versus-workflows/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#human-in-the-loop"},{"slug":"indirect-prompt-injection","term":"Indirect prompt injection","definition":"A prompt-injection attempt carried in external content such as a webpage or document rather than a direct user instruction.","source":"https://genai.owasp.org/llmrisk/llm01-prompt-injection/","learn":"https://loadedtokens.com/learn/mcp-and-tool-permissions/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#indirect-prompt-injection"},{"slug":"inference","term":"Inference","definition":"Using a trained model to compute an output from new input, rather than training its parameters.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/open-weights-versus-hosted/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#inference"},{"slug":"input-tokens","term":"Input tokens","definition":"Tokens processed as input to a model call; billing may distinguish cached input and other categories.","source":"https://developers.openai.com/api/docs/pricing","learn":"https://loadedtokens.com/learn/tokens-context-and-memory/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#input-tokens"},{"slug":"language-model","term":"Language model","definition":"A model that learns patterns in language and produces or scores token sequences. Its output is not automatically verified.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/what-a-language-model-does/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#language-model"},{"slug":"llm","term":"LLM","definition":"Large language model. A language model trained at substantial scale; “large” is not a standardized quality rating.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/what-a-language-model-does/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#llm"},{"slug":"mcp","term":"MCP","definition":"Model Context Protocol: a standard for connecting AI applications with tools, data and other capabilities.","source":"https://modelcontextprotocol.io/docs/getting-started/intro","learn":"https://loadedtokens.com/learn/mcp-and-tool-permissions/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#mcp"},{"slug":"mcp-client","term":"MCP client","definition":"The component that maintains a protocol connection to an MCP server on behalf of a host.","source":"https://modelcontextprotocol.io/docs/getting-started/intro","learn":"https://loadedtokens.com/learn/mcp-and-tool-permissions/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#mcp-client"},{"slug":"mcp-host","term":"MCP host","definition":"The AI application that manages clients and presents capabilities to the user.","source":"https://modelcontextprotocol.io/docs/getting-started/intro","learn":"https://loadedtokens.com/learn/mcp-and-tool-permissions/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#mcp-host"},{"slug":"mcp-server","term":"MCP server","definition":"A service exposing capabilities such as tools and resources through MCP. A connection is not authorization for every action.","source":"https://modelcontextprotocol.io/docs/getting-started/intro","learn":"https://loadedtokens.com/learn/mcp-and-tool-permissions/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#mcp-server"},{"slug":"model-card","term":"Model card","definition":"Documentation describing a model’s intended uses, training context and limitations; details vary by publisher.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/open-weights-versus-hosted/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#model-card"},{"slug":"model-weights","term":"Model weights","definition":"The learned numerical parameters of a model. Access to them is distinct from the license governing their use.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/open-weights-versus-hosted/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#model-weights"},{"slug":"multimodal","term":"Multimodal","definition":"Able to process or generate more than one kind of data, such as text and images. Supported combinations depend on the model.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/choose-a-model-by-the-job/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#multimodal"},{"slug":"open-weights","term":"Open weights","definition":"A description of model weights being available to obtain. It does not by itself promise unrestricted use or open training data.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/open-weights-versus-hosted/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#open-weights"},{"slug":"orchestration","term":"Orchestration","definition":"Coordinating steps, state, tools and decisions across a workflow or agent system.","source":"https://docs.langchain.com/oss/python/langgraph/overview","learn":"https://loadedtokens.com/learn/agents-versus-workflows/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#orchestration"},{"slug":"output-tokens","term":"Output tokens","definition":"Tokens generated by a model call. Their price can differ from input token pricing.","source":"https://developers.openai.com/api/docs/pricing","learn":"https://loadedtokens.com/learn/tokens-context-and-memory/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#output-tokens"},{"slug":"prompt","term":"Prompt","definition":"The instructions and input supplied to a model for a task. Relevant context and examples may be part of it.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/prompting-with-a-brief/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#prompt"},{"slug":"prompt-caching","term":"Prompt caching","definition":"Reusing eligible repeated context under a provider’s caching rules. Eligibility, retention and price are provider-specific.","source":"https://developers.openai.com/api/docs/pricing","learn":"https://loadedtokens.com/learn/tokens-context-and-memory/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#prompt-caching"},{"slug":"prompt-injection","term":"Prompt injection","definition":"An attempt to redirect model behavior using instructions in input, including external material the application retrieves.","source":"https://genai.owasp.org/llmrisk/llm01-prompt-injection/","learn":"https://loadedtokens.com/learn/mcp-and-tool-permissions/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#prompt-injection"},{"slug":"rag","term":"RAG","definition":"Retrieval-augmented generation: retrieve relevant material, then supply it as context when generating an answer.","source":"https://platform.openai.com/docs/guides/embeddings","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#rag"},{"slug":"retrieval","term":"Retrieval","definition":"Finding candidate evidence or records for a task. A retrieval error can give a model the wrong basis for its answer.","source":"https://platform.openai.com/docs/guides/embeddings","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#retrieval"},{"slug":"risk-management","term":"Risk management","definition":"Identifying, measuring and managing risks in context throughout a system’s lifecycle rather than relying on a single safety claim.","source":"https://www.nist.gov/itl/ai-risk-management-framework","learn":"https://loadedtokens.com/learn/safe-workplace-ai/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#risk-management"},{"slug":"rubric","term":"Rubric","definition":"Explicit grading criteria defining an acceptable output and the severity of failures.","source":"https://platform.openai.com/docs/guides/evals","learn":"https://loadedtokens.com/learn/evals-and-cost-per-success/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#rubric"},{"slug":"semantic-search","term":"Semantic search","definition":"Search using meaning-related representations rather than relying only on exact word matches.","source":"https://platform.openai.com/docs/guides/embeddings","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#semantic-search"},{"slug":"state","term":"State","definition":"Information retained across workflow steps, such as intermediate results and pending approvals.","source":"https://docs.langchain.com/oss/python/langgraph/overview","learn":"https://loadedtokens.com/learn/agents-versus-workflows/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#state"},{"slug":"structured-output","term":"Structured output","definition":"An output constrained to a defined structure, such as a schema. Matching a schema does not establish factual accuracy.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/prompting-with-a-brief/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#structured-output"},{"slug":"system-instruction","term":"System instruction","definition":"An instruction supplied by the application to guide model behavior, distinct from ordinary task data. Exact roles vary by API.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/prompting-with-a-brief/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#system-instruction"},{"slug":"token","term":"Token","definition":"A unit a model processes, often a word fragment or punctuation in text. A token is not always a word.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/tokens-context-and-memory/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#token"},{"slug":"tokenization","term":"Tokenization","definition":"Converting input into model-compatible token units. Different tokenizers can split the same text differently.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/tokens-context-and-memory/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#tokenization"},{"slug":"tool-call","term":"Tool call","definition":"A request from a model-driven application to an external function or capability. The application controls execution and permission.","source":"https://www.anthropic.com/engineering/building-effective-agents","learn":"https://loadedtokens.com/learn/mcp-and-tool-permissions/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#tool-call"},{"slug":"trace","term":"Trace","definition":"A record of execution steps and intermediate activity used for debugging or inspection. It may contain sensitive content.","source":"https://docs.langchain.com/oss/python/langgraph/overview","learn":"https://loadedtokens.com/learn/evals-and-cost-per-success/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#trace"},{"slug":"training","term":"Training","definition":"The process of adjusting model parameters using data and an optimization objective. It differs from adding context to a prompt.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/what-a-language-model-does/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#training"},{"slug":"transformer","term":"Transformer","definition":"A neural-network architecture widely used in modern language models, with attention mechanisms for processing relationships in sequences.","source":"https://huggingface.co/learn/llm-course/en/chapter1/1","learn":"https://loadedtokens.com/learn/what-a-language-model-does/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#transformer"},{"slug":"vector-search","term":"Vector search","definition":"Retrieving items by similarity between numerical representations. Similarity does not prove truth or freshness.","source":"https://platform.openai.com/docs/guides/embeddings","learn":"https://loadedtokens.com/learn/rag-without-the-magic/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#vector-search"},{"slug":"workflow","term":"Workflow","definition":"A sequence of steps with a predefined route. It may contain model calls without becoming fully autonomous.","source":"https://www.anthropic.com/engineering/building-effective-agents","learn":"https://loadedtokens.com/learn/agents-versus-workflows/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#workflow"},{"slug":"zero-shot-prompting","term":"Zero-shot prompting","definition":"Requesting a task without worked examples in that prompt. It does not mean the model had no training.","source":"https://platform.openai.com/docs/guides/prompt-engineering","learn":"https://loadedtokens.com/learn/prompting-with-a-brief/","updated":"2026-11-07","url":"https://loadedtokens.com/learn/glossary/#zero-shot-prompting"}]}