constitutional-ai

RL study guide — foundations through RLHF, DPO, GRPO, RLVR, agentic RL, and offline RL. Hand-written CS294 notes, 19 lecture drafts, 5 tested exercises, citations that resolve.

Smart Promise Tracker AI 2026 – Commit Memory Engine

[L0 CONSTITUTION] arifOS — constitutional MCP kernel. Law, identity, F1–F13, VAULT999. Judges but never executes. DITEMPA BUKAN DIBERI.

CORE is a governance runtime for autonomous AI systems. It enforces constitutional rules during execution, prevents governance bypass, and creates auditable authority chains for agent actions across operational domains.

一份来自2026年的AI精神病理学诊断报告 / A Pathological Diagnosis of AI Civilization

A framework for healthy human-agent collaboration. Tell your AI coding agent when to stop helping you.

An autonomous LLM research agent that executes the full academic paper pipeline — from literature search to compiled PDF

Glass Box Framework — runtime constitutional verification for AI answers. Trust Cards with claim-level reasoning chains, formal ECS scoring, the 7-angle Glassbox Court red team, and deterministic audit logs. MCP-native.

A production-grade LLM architecture built from scratch in PyTorch. Features Multi-Head Latent Attention (MLA), Mixture of Experts (MoE), GRPO alignment, and a complete 31-part educational course.

ARI — Artificial Reasoning Intelligence. Personal AI operating system with 7-layer architecture, three-pillar cognition (LOGOS/ETHOS/PATHOS), constitutional governance, and tamper-evident SHA-256 audit trails. Local-first. Multi-agent. TypeScript.

Production-ready AI multi-agent orchestration platform. 37+ modules, constitutional governance, RAG knowledge integration, real-time collaboration. Kubernetes & Docker ready.

ASAN: A conceptual architecture for a self-creating (autopoietic), energy-efficient, and governable multi-agent AI system.

Constitutional AI platform where everyone gets an Angel. Multi-tenant marketplace with LEO AI, SSE streaming, vendor onboarding, product configurator, reviews, ultimate fair split. Payload CMS 3.77 + Next.js 16 + React 19 + PostgreSQL.

Constitutional framework for human sovereignty in the AI-native era. A systematic approach to preserving human agency through 23 core protocols and the Dual-Circle Symbiosis Model.

A complete breakdown of the Fable 5 constitutional AI jailbreak technique for Claude 4.8

Constitutional midtraining for LLM alignment: value-corpus generation, 2×2 factorial training pipeline, and full evaluation suite (120B-parameter model)

An experiment, not a tool: an autonomous agent on a local LLM that proposes changes to its own constitution and values; a human approves every one. Value layer (constitution, identity, skills, rules) sits behind a human approval gate, every change replayable. Runs on Ollama, holds up on a 16 GB Mac, no cloud LLM. Security by absence.

CORD — Constitutional AI safety engine for autonomous agents. Hard blocks, risk scoring, real-time audit. 482 tests.

Bench is a constitutional governance layer for Claude Code

Constitutional governance membrane between agent reasoning and side effects. No valid Decision Receipt, no side effect. Apache-2.0, Beta. Not an agent framework.