adversarial-validation

A full pipeline AutoML tool for tabular data

A tiny framework to perform adversarial validation of your training and test data.

A 4-stage adversarial research auditor that fetches papers from arXiv & HuggingFace, extracts claims, and uses DeepSeek-R1 to verify them against raw abstracts.A self-correcting research & paper digest pipeline powered by local LLMs & reasoning agents

Distributed Collection, Local Intelligence - Stop LLMs from hallucinating with Kong in the Loop architecture

Turn LLM priors into scientific rigor. Zero-drift multi-agent framework for reproducible research code.

Builds a fraud detection system on IEEE-CIS data that explicitly models temporal distribution shift by training adversarial validators to detect when the production distribution diverges from training data, then dynamically reweights ensemble members (LightGBM, CatBoost, XGBoost) based on their robustness to detected drift regimes.

The AI coding partner that doesn't trust itself. Local inline code hints, adversarial skeptic validation, full session replays, and dynamic GPU auto-scaling.

Grounded answers, control mapping, and audit-ready exports from bounded source material.

Covariate-shift correction by adversarial-validation / propensity weighting toward the target distribution. Honest: helps under misspecification+shift, ties null when well-specified; never uses target y. Held-out + null validated. numpy-only.

CCCE: Adversarial Validation Framework for Quantum Circuit Optimization