A daily crypto, finance, and AI vocabulary grouping puzzle covering backing checks, lending risk, model serving, and retrieval workflows.
Stablecoin Reserve Signals
These words describe what supports a dollar-linked token and how holders check the peg.
- Reserve Ratio: The reserve ratio compares backing assets with the liabilities they are meant to support.
- Attestation: An attestation is a report that checks an issuer's balances or reserves at a point in time.
- Redemption Window: A redemption window is the time or process a holder uses to convert a token back into its reference asset.
- Peg Band: A peg band is a small range around a target price that helps show whether a token is near its anchor.
DeFi Credit Risk
These words explain how overcollateralized lending keeps accounts healthy until prices move.
- Collateral Ratio: The collateral ratio compares the value of pledged assets with the debt they secure.
- Health Factor: A health factor is a risk score that shows how far a borrowing position is from liquidation.
- Borrow Rate: A borrow rate is the interest charged when a user takes out a loan against collateral.
- Liquidation Threshold: The liquidation threshold is the level at which a position can be closed because collateral is no longer sufficient.
AI Inference Stack
These words cover the runtime pieces that shape speed, cost, and throughput when a model answers.
- Inference: Inference is the process of using a trained model to produce predictions or answers.
- Batching: Batching groups multiple requests so a model can process them together more efficiently.
- Quantization: Quantization reduces numeric precision to make model storage and inference more efficient.
- Throughput: Throughput is the amount of work a system can process in a given period.
Retrieval Data Ops
These words describe how applications store facts, search them, and keep them current.
ChunkingEmbeddingVector StoreReranker
- Chunking: Chunking breaks content into smaller pieces so it can be indexed or retrieved more easily.
- Embedding: An embedding is a numeric representation that lets a model compare meaning across items.
- Vector Store: A vector store is a database for saving and searching embeddings.
- Reranker: A reranker is a model that reorders retrieved results by relevance before they are shown to a user or agent.