Toolkit to assess and determine model provenance
Python toolkit for fingerprinting anonymous LLMs on OpenAI-compatible endpoints using tokenizer, context, error, and vision probes
🔍 Detect API model dilution and fingerprint which LLM is actually running behind any OpenAI-compatible or Anthropic endpoint
Codex GPT 模型指纹检测：自动采样、本地判分、结果记录与一键重测。
LLM relay reality check — a CLI that audits whether an API relay actually serves the model it claims (货不对板检测)。
Smudged Fingerprints: evaluating AI image attribution under fingerprint removal and forgery attacks (IEEE SaTML 2026).
AI 中转站 API 验真取证工具：模型验真 · 系统提示词提取 · 上下游供应链溯源 | LLM relay API forensics: verify models, extract system prompts, trace upstream supply chains
Open-source implementation and canonical pipelines for SIF-VLM fingerprint research
CISPA European Championship 2026 — 3-task ML security hackathon: black-box image reconstruction, generative model fingerprinting, and hardware-dependent chimera generation.
A multi-layer verification system for AI-generated analysis that exploits documented behavioral differences across language models to detect hallucination, simulation drift, and epistemic failures.
Open benchmark for OML fingerprint robustness across on-device model conversion and quantization
Passive LLM endpoint identity and stability middleware for Python and Node.js
Official implementation for Efficient and Robust DNN Model Fingerprint with Decision Difference Regions Identification (Applied Intelligence).
Detect what model is really behind any OpenAI-compatible proxy, reseller or gateway. 6-way behavioral forensics - tokenizer fingerprint, trap-question bank, speed side channel, tool-call integrity - with no self-declaration needed. Zero-dependency single binary (Win/Linux/macOS), exit codes built for CI.
CLI framework for auditing LLM providers and detecting model substitution.
Adversarial token convergence experiments on MLX. Pits language models against each other to reveal training strata — solo mode converges in 2 iterations, adversarial mode never does. Model fingerprinting through pressure.
Local model-route receipts for ChatGPT and Codex. Inspect the evidence, compare requests, and share what your client reported.
Local-first LLM API integrity checker with statistical fingerprinting and optional EvalScope capability evaluation
To associate your repository with the model-fingerprinting topic, visit your repo's landing page and select "manage topics."