Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 
 
 
 
 
 
 

README.md

ModelAudit

AI model performance verification with auditor staking and benchmark challenges on ProbeChain Rydberg Testnet.

Features

  • Model Registration — Register AI models across categories (NLP, Vision, Audio, Multimodal, Tabular, Reinforcement)
  • Auditor Staking — Auditors stake tokens to submit verified benchmarks
  • Benchmark Submission — Submit scored benchmarks with proof hashes
  • Challenge System — Challenge suspicious results with stake-backed disputes
  • Slashing — Auditors lose stake when challenges are upheld

Contract: ModelBenchmark.sol

Function Description
registerModel(name, modelHash, category) Register a model
stakeAsAuditor() Stake to become an auditor
submitBenchmark(modelId, benchmarkId, score, proofHash) Submit benchmark
getModelScore(modelId) Get average model score
challengeResult(benchmarkId, reason) Challenge a result

Setup

npm install
cp .env.example .env
# Edit .env with your private key
npx hardhat compile
npx hardhat run scripts/deploy.ts --network rydberg

Network