A daily crypto, finance, and AI vocabulary puzzle covering settlement rails, collateral risk, search pipelines, and model serving.
Stablecoin Controls
These words describe how a stablecoin stays near its peg through reserves, disclosures, and redemption mechanics.
- Reserve Backing: Reserve backing is the pool of assets that supports a stablecoin's claimed value.
- Attestation: An attestation is a signed statement that reserves, balances, or data have been verified.
- Redemption Window: A redemption window is the period when holders can exchange a stablecoin for its backing asset.
- Peg Band: A peg band is the price range a stablecoin can move within before market stress becomes visible.
Credit Risk
These words describe leverage limits, collateral coverage, and when a position can be liquidated.
- Collateral Factor: Collateral factor is the share of posted collateral that can count toward borrowing capacity.
- Loan-to-Value: Loan-to-value compares the size of a loan with the value of the collateral securing it.
- Health Factor: Health factor is a risk score that shows how close a borrowing position is to liquidation.
- Liquidation Threshold: Liquidation threshold is the collateral level at which a position can be forcibly closed.
Retrieval Systems
These words describe how embeddings, search indexes, and ranking stages help a model find relevant context.
- Embedding: An embedding is a vector that encodes text, images, or other data for similarity search.
- Vector Store: A vector store stores embeddings so an app can retrieve nearby items quickly.
- Retrieval: Retrieval is the step where a system finds documents or chunks relevant to a query.
- Reranker: A reranker is a model that reorders retrieved results to improve relevance.
Inference Ops
These words describe serving tradeoffs such as memory reuse, precision, and throughput.
- Batching: Batching combines multiple inference requests so they can run more efficiently together.
- KV Cache: A KV cache stores key and value tensors so a model can reuse attention history.
- Quantization: Quantization reduces numerical precision to make model serving faster and lighter.
- Throughput: Throughput measures how many requests or tokens a serving system can process over time.