bias_amplification_tracker
ActiveTool of gapup-mcp
declared in 0.2.0
Tracks bias amplification in LLM outputs by analyzing fairness metrics from HuggingFace's model leaderboard. Designed for risk assessment personas to detect and quantify demographic, gender, or racial bias amplification in generated text. Accepts model identifiers or output samples, returns structured bias metrics and amplification trends.
Parameters schema
{
"type": "object",
"required": [],
"properties": {
"async": {
"type": "boolean",
"description": "If true, returns a job_id immediately (<200ms) instead of waiting for the result. Poll the result with job_result(job_id). Use for slow tools to avoid client timeouts."
},
"modelId": {
"type": "string",
"description": "HuggingFace model identifier (e.g., 'facebook/opt-1.3b')"
},
"outputSamples": {
"type": "array",
"items": {
"type": "string"
},
"description": "Array of LLM output strings to analyze for bias amplification"
},
"demographicGroups": {
"type": "array",
"items": {
"type": "string"
},
"description": "Specific demographic groups to monitor (e.g., ['gender', 'race'])"
}
}
}No endpoints wrapped at confidence ≥ 0.70.
Parent server
gapup-mcp
https://github.com/getgapup/gapup-mcp-public
2/7 registries