surface↓classification↓audience

Classificationdimension · returns varchar

AUDIENCE

Identify target audience via zero-shot NLI

Per-row classifier — stable across GROUP BY.

classificationnlispecialist-zootextscales-large

Arguments

name	type	description
text	VARCHAR	—
domain	VARCHAR	—

About

Tag each piece of text by who it's written for — the intended reader. Handy when you have a mixed-audience content library (a blog, a help center, a support archive) and want to segment it by audience tier before further analysis, routing, or reporting. Returns one of a fixed set of audience labels such as general public, experts, beginners, or practitioners. Pass an optional `domain` argument to steer the classification toward a specific field — for example, `domain='tech'` nudges the labels toward technical-audience tiers. Fast and cheap: runs a zero-shot model on the specialist zoo GPU rather than calling a full LLM, so it scales across large corpora without blowing up your token budget. If your label set is custom or the domain requires contextual judgment beyond the fixed bucket list, reach for AUDIENCE_LLM instead.

Examples

Audience dimension returned

SELECT
  audience ('Scientific paper on quantum mechanics')

Nearby rabbit holes

same domain

AUDIENCE_LLM

LLM-backed audience identification (escape hatch for AUDIENCE)

classificationllmllm-escape-hatchtext

AUTHENTICITY

Assess authenticity of content via zero-shot NLI

classificationnlispecialist-zootext

AUTHENTICITY_LLM

LLM-backed authenticity assessment (escape hatch for AUTHENTICITY)

classificationllmllm-escape-hatchtext

BUCKET

Classify text into user-specified buckets via zero-shot NLI

classificationnlispecialist-zootext

BUCKET_LLM

LLM-backed bucketing (escape hatch for BUCKET)

classificationllmllm-escape-hatchtext

CASE

Multi-branch semantic classification via specialist zoo NLI

classificationnlispecialist-zootext

Climb back to The Looking Glass

surface↓classification↓audience

Classificationdimension · returns varchar

AUDIENCE

Identify target audience via zero-shot NLI

Per-row classifier — stable across GROUP BY.

classificationnlispecialist-zootextscales-large

Arguments

name	type	description
text	VARCHAR	—
domain	VARCHAR	—

About

Tag each piece of text by who it's written for — the intended reader. Handy when you have a mixed-audience content library (a blog, a help center, a support archive) and want to segment it by audience tier before further analysis, routing, or reporting. Returns one of a fixed set of audience labels such as general public, experts, beginners, or practitioners. Pass an optional `domain` argument to steer the classification toward a specific field — for example, `domain='tech'` nudges the labels toward technical-audience tiers. Fast and cheap: runs a zero-shot model on the specialist zoo GPU rather than calling a full LLM, so it scales across large corpora without blowing up your token budget. If your label set is custom or the domain requires contextual judgment beyond the fixed bucket list, reach for AUDIENCE_LLM instead.

Examples

Audience dimension returned

SELECT
  audience ('Scientific paper on quantum mechanics')

Nearby rabbit holes

same domain

AUDIENCE_LLM

LLM-backed audience identification (escape hatch for AUDIENCE)

classificationllmllm-escape-hatchtext

AUTHENTICITY

Assess authenticity of content via zero-shot NLI

classificationnlispecialist-zootext

AUTHENTICITY_LLM

LLM-backed authenticity assessment (escape hatch for AUTHENTICITY)

classificationllmllm-escape-hatchtext

BUCKET

Classify text into user-specified buckets via zero-shot NLI

classificationnlispecialist-zootext

BUCKET_LLM

LLM-backed bucketing (escape hatch for BUCKET)

classificationllmllm-escape-hatchtext

CASE

Multi-branch semantic classification via specialist zoo NLI

classificationnlispecialist-zootext

Climb back to The Looking Glass