surface↓classification↓credibility

Classificationdimension · returns varchar

CREDIBILITY

Analyze credibility/reliability of text descriptions (zero-shot NLI)

Per-row classifier — stable across GROUP BY.

classificationnlispecialist-zootext

Arguments

name	type	description
text	VARCHAR	—
strictness	VARCHAR	—
num_levels	INTEGER	—

About

Semantic credibility dimension — analyzes reliability/trustworthiness of text descriptions or claims. Dimension-shaped for GROUP BY clauses. Backend: specialist zoo zero-shot NLI via /classify_batch. Every text in the batch is scored against a fixed credibility scale in a single forward pass. Fixed label set (5 levels): Very Low Credibility / Low Credibility / Moderate Credibility / High Credibility / Very High Credibility The optional `strictness` argument (lenient/balanced/strict) shifts the hypothesis template to bias the model toward that judgment style — strict mode uses "This text provides strong verifiable evidence and specific detail: {label}", lenient uses softer phrasing. For custom strictness rules, use CREDIBILITY_LLM.

Examples

Credibility assessed for well-sourced claim

SELECT
  credibility (
    'Peer-reviewed study shows 87% confidence interval on treatment efficacy with n=2000'
  )

Nearby rabbit holes

same domain

AUDIENCE

Identify target audience via zero-shot NLI

classificationnlispecialist-zootext+1

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

Climb back to The Looking Glass

surface↓classification↓credibility

Classificationdimension · returns varchar

CREDIBILITY

Analyze credibility/reliability of text descriptions (zero-shot NLI)

Per-row classifier — stable across GROUP BY.

classificationnlispecialist-zootext

Arguments

name	type	description
text	VARCHAR	—
strictness	VARCHAR	—
num_levels	INTEGER	—

About

Semantic credibility dimension — analyzes reliability/trustworthiness of text descriptions or claims. Dimension-shaped for GROUP BY clauses. Backend: specialist zoo zero-shot NLI via /classify_batch. Every text in the batch is scored against a fixed credibility scale in a single forward pass. Fixed label set (5 levels): Very Low Credibility / Low Credibility / Moderate Credibility / High Credibility / Very High Credibility The optional `strictness` argument (lenient/balanced/strict) shifts the hypothesis template to bias the model toward that judgment style — strict mode uses "This text provides strong verifiable evidence and specific detail: {label}", lenient uses softer phrasing. For custom strictness rules, use CREDIBILITY_LLM.

Examples

Credibility assessed for well-sourced claim

SELECT
  credibility (
    'Peer-reviewed study shows 87% confidence interval on treatment efficacy with n=2000'
  )

Nearby rabbit holes

same domain

AUDIENCE

Identify target audience via zero-shot NLI

classificationnlispecialist-zootext+1

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

Climb back to The Looking Glass