new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

jina-ai/options

6 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

Jina provider options: inputType

The inputType provider option specifies the intended downstream application to help the model produce better embeddings. Type: 'text-matching' | 'retrieval.query' | 'retrieval.passage' | 'separation' | 'classification'. Default: 'retrieval.passage'. Valid values: 'retrieval.query' (input is a search query), 'retrieval.passage' (input is a document/passage), 'text-matching' (for semantic textual similarity tasks), 'classification' (for classification tasks), 'separation' (for clustering tasks).

Jina provider options: outputDimension

The outputDimension provider option specifies the number of dimensions for the output embeddings. Type: number. Valid ranges depend on the model: jina-embeddings-v3 min 32, max 1024. jina-clip-v2 min 64, max 1024. jina-clip-v1 fixed 768.

Jina provider options: embeddingType

The embeddingType provider option specifies the data type for the returned embeddings. Type: 'float' | 'binary' | 'ubinary' | 'base64'.

Jina provider options: normalized

The normalized provider option specifies whether to L2-normalize embeddings. Type: boolean. Default: true.

Jina provider options: truncate

The truncate provider option specifies whether to truncate inputs beyond the model context limit instead of erroring. Type: boolean. Default: false.

Jina provider options: lateChunking

The lateChunking provider option specifies whether to split long inputs into 1024-token chunks automatically. Type: boolean. Only available for text embedding models.

Give your agent this brain