accountability

Examples of techniques for training interpretable ML models, explaining ML models, and debugging ML models for accuracy, discrimination, and security.

H2O.ai Machine Learning Interpretability Resources

Personal AI assistant for work inside corporate constraints, built on coding agents and the tools, sessions, and permissions you already have.

A library that implements fairness-aware machine learning algorithms

The first AI plugin that speaks first. Code-enforced learning + active forgetting + PAC (Proactive Accountability Challenge). Works with Claude Code, Gemini CLI, Hermes, OpenClaw.

Modular Python Toolbox for Fairness, Accountability and Transparency Forensics

A national archive of police data collected by journalists, lawyers, and activists around the country.

youtube & tiktok analysis + youchoose recommendation custmizer. backend, extensions, and tooling

`GitHub` `Python` `PDF` `Public-records` - Public-records archive and research tools for the DOJ Epstein files, with document indexes, manifests and visualizations. https://ringmast4r.github.io/Epstein/flights.html

The OpenPoliceData (OPD) Python library is the most comprehensive centralized public access point for incident-level police data in the United States. OPD provides easy access to 550+ incident-level datasets from 236 police agencies and 11 entire states. Types of data include traffic stops, use of force, officer-involved shootings, and complaints.

Track Indian Parliamentary Committee reports from sansad.in

A neuro-symbolic runtime architecture (DIR) for Responsibility-Oriented Agents (ROA). By decoupling probabilistic LLM reasoning (User Space) from deterministic execution (Kernel Space), it ensures safe, invariant-driven, and auditable AI autonomy.

A curated list of tools and resources for online adult content/porn filtering

Prior direction, kept rather than deleted. Signed, offline-verifiable receipts for AI agent actions, and a reference implementation of the OWASP Agentic Skills Top 10 AST09 receipt pattern. Nobulex is now the independent reliability registry for agent tools: github.com/arian-gogani/nobulex-registry

Project Lockdown (an initiative from The IO Foundation) is a civic tech, interactive platform providing an overview of the state of Human and Digital Rights around the globe. It evaluates policies obtained from official sources that may impact their observance. It provides, among other tools, a layered map interface that allows for a visual repr…

Proof of Understanding. Starting with GitHub: show that you understand the change before it merges.

SenateTrades: what stocks are your senators buying?

A Python wrapper for the OpenFEC API.

A cognition kernel that wraps a frozen LLM in a persistent identity + trust + governance loop — testing whether capability comes from structure, not weights. Research-stage; honest about what's real vs mocked.

Chilean Municipalities Information System (SINIM) Wrapper 📈🏛🇨🇱