AIIntermediate#AI Safety#Data Provenance#Limitations
Context Window
C
Definition
The maximum input and output token capacity a model can handle in one request, subject to the model and provider configuration. It is a capacity limit, not dependable memory, data verification, or a measure of reasoning quality.
Source reviewed
- NIST: Adversarial Machine Learning Taxonomy opens in a new tab
Supports: NIST identifies indirect prompt injection, data poisoning, and other integrity risks in generative-AI systems that process untrusted input.
- OWASP: LLM01 Prompt Injection opens in a new tab
Supports: OWASP documents that user or retrieved content can alter an LLM application's behavior and recommends keeping sensitive functions outside model-controlled instructions.
Concept Diagram
Key Takeaways
1
The context window is a request capacity for tokenized input and output, not reliable memory or a quality score.
2
For every included record, retain its source, time, identifier, transformation, and whether it was omitted, truncated, or summarized.
3
Test for attention, retrieval, conflict, malformed-content, and prompt-injection failures; compare critical output with primary records.
4
More capacity can increase cost and latency without improving factual accuracy or authorizing a state-changing action.
Practical Example
A research workflow sends an official protocol notice, its original URL, timestamps, and a small set of labelled related records to a model. It reports which record was truncated and asks for a source-linked summary. It does not infer a price move or create a transaction from the context.
Finished learning?
Related Terms
AI
Perception LayerThe part of an AI workflow that collects and labels inputs such as blockchain records, market data, and public communications. It supplies context for review; it does not establish truth, identify people, predict prices, or justify an action.
AI
LLM ReasoningUsing a language model to summarize, classify, or compare supplied information. Its output is probabilistic text, not verified market analysis, financial advice, or authority to execute a trade.
AI
Prompt EngineeringThe practice of designing model instructions, examples, context, and output constraints for a bounded task. It can make a workflow easier to evaluate; it does not make model output deterministic, correct, secure, or suitable for a financial decision.
AI
RAG (Retrieval-Augmented Generation)A system pattern that retrieves selected external records and supplies them with a query to a language model. Retrieval can improve traceability and recency when sources are cited; it does not prove that a record is true, complete, current, or safe to act on.
Related Deep Dives
Master the hub of AI
Explore all our strategic guides about AI to take your operations to the next level.
View all articles