惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

博客园_首页
阮一峰的网络日志
阮一峰的网络日志
S
Secure Thoughts
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
C
Cyber Attacks, Cyber Crime and Cyber Security
T
Threat Research - Cisco Blogs
P
Privacy & Cybersecurity Law Blog
The Hacker News
The Hacker News
H
Heimdal Security Blog
W
WeLiveSecurity
L
LINUX DO - 热门话题
Hacker News: Ask HN
Hacker News: Ask HN
WordPress大学
WordPress大学
The Last Watchdog
The Last Watchdog
Hugging Face - Blog
Hugging Face - Blog
博客园 - 【当耐特】
D
DataBreaches.Net
I
Intezer
Webroot Blog
Webroot Blog
C
Cisco Blogs
AWS News Blog
AWS News Blog
博客园 - 聂微东
T
The Blog of Author Tim Ferriss
V
Vulnerabilities – Threatpost
罗磊的独立博客
Google DeepMind News
Google DeepMind News
N
Netflix TechBlog - Medium
Schneier on Security
Schneier on Security
宝玉的分享
宝玉的分享
博客园 - 叶小钗
PCI Perspectives
PCI Perspectives
D
Docker
Scott Helme
Scott Helme
NISL@THU
NISL@THU
J
Java Code Geeks
B
Blog RSS Feed
Google Online Security Blog
Google Online Security Blog
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
T
The Exploit Database - CXSecurity.com
AI
AI
美团技术团队
Cloudbric
Cloudbric
月光博客
月光博客
P
Proofpoint News Feed
T
Tailwind CSS Blog
Google DeepMind News
Google DeepMind News
小众软件
小众软件
www.infosecurity-magazine.com
www.infosecurity-magazine.com
The Cloudflare Blog
cs.CV updates on arXiv.org
cs.CV updates on arXiv.org

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant Common SOC 2 Failures (Real World) Stop Vibe-Checking Your AI App: A Practical Guide to Evals How to Use SonarQube and SonarScanner Locally to Level Up Your Code Quality Your Next To-Do App Is Dead — I Replaced Mine with an OpenClaw AI Sign a Nostr event in 60 lines of Python using coincurve — no nostr-sdk, no nbxplorer, no rust toolchain ITGC Audit Explained Like You’re in Big 4 Patch Tuesday abril 2026: Microsoft parcha 163 vulnerabilidades y un zero-day en SharePoint Stop scraping everything: a better way to track competitor price changes Listing on MCPize + the Official MCP Registry while routing payments OUTSIDE the marketplace — how I kept 100% of my x402 revenue Building an AI-Powered Risk Intelligence System Using Serverless Architecture Why We Ripped Function Overloading Out of Our AI Toolchain Testing AI-Generated Code: How to Actually Know If It Works SaaS Churn Is Killing Your Business. Here Is What to Do About It (Without a Support Team) The Speed of AI Is No Longer Linear - And Self-Improving Models Are Why How to Implement RBAC for MCP Tools: A Practical Guide for Engineering Teams From Standard Quote to Persuasive Proposal: AI Automation for Arborists I built a CLI that scaffolds complete multi-tenant SaaS apps Axios CVE-2025–62718: The Silent SSRF Bug That Could Be Hiding in Your Node.js App Right Now The dashboard that ended our friendship Data Pipelines Explained Simply (and How to Build Them with Python) The Hidden Cost of AI Systems Nobody Talks About. undefined vs undeclared, and how typeof behaves Switching from file-based jobs to NATS/Kafka in Rust without changing code io_uring Adventures: Rust Servers That Love Syscalls Why Agentic AI is Killing the Traditional Database The POUR principles of web accessibility for developers and designers Quantum Neural Network 3D — A Deep Dive into Interactive WebGL Visualization How To Install Caveman In Codex On macOS And Windows Automation Pipeline Reliability: Why Your Workflow Breaks When Nobody Is Watching I Built an 'Open World' AI Coding Agent — It Works From ANY Folder From Freelancing to Product: A Tech Service Company's SaaS Transformation China's AI Giants: Adding Tencent Hunyuan & ByteDance Doubao to AI University (74 Providers) On the Vibe Coders and Their Lies clerk: Auto-Summarize Your Claude Code Sessions AI Weekly — 2026/04/10–04/17 | The Model Lockdown Is Here, but the Toolchain Is the Real Battleground AI 週報 — 2026/04/10–2026/04/17 模型封鎖潮來了,但工具鏈才是真戰場 Maybe this is how Open-Source apps are born... 🚀 Fine-Tune LLMs with LoRA and QLoRA: 2026 Guide tRPC v11 + Next.js App Router: End-to-End Type Safety Without the Boilerplate ShadCN UI in 2026: Why I Stopped Installing Component Libraries and Started Owning My Components SaaS Billing in React Server Components: Stripe + Supabase Without a Single `useEffect` Join our DEV Weekend Challenge — $1,000 in Prizes Across TEN winners! Submissions Due April 20 at 6:59 AM UTC. Implementing FSRS Spaced Repetition in Flutter + Supabase — Adding Memory Science to an AI Learning App "I Texted My Localhost From the Train — Claude Code Fixed the Bug Before I Got Home" I Built a Sales Prep AI and It Went Deeper Than Expected Design to Code #2: One JSON, Eleven Outputs Solving the 100M-Row Problem: A Summary Table Pattern for High-Volume Push Notification Logs Flutter Web With Wasm: What Actually Changes For Developers I Built 50 Royalty-Free Soundtracks for My Side Project in a Weekend Using AI Music Generation The Vibe Coding Security Checklist: 7 Things to Check Before You Ship Stop Letting Googlebot Guess Fix Your React App's SEO Right Desconstruindo o Streaming do LinkedIn: Como Criar um Engine de Extração de Vídeo de Alta Performance com HLS e FFmpeg (EDA Part-1) EDA (Exploratory Data Analysis) Explained With Real Life — Why Looking at Your Data Is the Most Important Step in Machine Learning Brand Relationship Management at Scale: Our 4-Touch Outreach System for 200+ Brands Why String.fromEnvironment() Might Return an Empty String in Dart JGuardrails 1.0.0 — Hardening Java LLM Apps Against Jailbreaks, Toxicity, and Prompt Injection Plan and Schedule a Full Week of Threads Content From One Claude Conversation Coding Cat Oran Ep3, Five Tables Changed Everything Updated: BFF Pattern I'm done watching freelancers get buried by 200 proposals. So I'm building the alternative. This is my first post BFS Algorithm in Java Step by Step Tutorial with Examples Tracking LLM Pricing Monthly: An Open Dataset for 22 AI Models How We Measure Content ROI on a Comparison Site: Revenue Attribution Without Perfect Data Introducing Nova AI Ops: The AI-Native Operating System for SRE Teams I built a free desktop video downloader for Windows — Grabbit How Talkie OCR Helps Vision-Impaired & Dyslexic Users Read the World Around Them VRCFaceTracking安装和iPhone面捕配置教程,有bug Even CrowdStrike Can't See Your Agents The Automation Gold Rush: What n8n Workflows and Claude Are Opening Up for Developers Right Now
Anti Refusal LLM Service
Taylor · 2026-05-31 · via DEV Community

I Built a 12MB Desktop App for Running Uncensored AI Models Locally (Tauri + Rust + Ollama) published: true description: How I built Cerberus AI — a local-first desktop app that auto-detects your GPU, pulls the right model quantization, and gives you uncensored AI chat without sending a single prompt to the cloud. Every major language model ships with an alignment layer that refuses certain prompts. Sometimes that's reasonable. Sometimes you're a security researcher, a creative writer, or just someone who doesn't want a corporation deciding what questions you're allowed to ask.

I built Cerberus AI to fix that — and to make the whole experience local-first, lightweight, and dead simple to install.

What Is Cerberus AI?
Cerberus AI is a platform for running open-weight, refusal-ablated language models on your own hardware. It has three parts:

A native desktop app (~12 MB) built with Tauri + Rust — not Electron
Open-weight GGUF models hosted on a public CDN
An OpenAI-compatible managed API for when you don't want to run local
The desktop app integrates directly with Ollama, auto-detects your GPU VRAM, and recommends the right model quantization for your hardware. From 4 GB laptops to 24 GB workstations, it just works.

Cerberus AI Desktop Chat

What Is Refusal Ablation?
This is the core technical innovation behind Cerberus models. Here's the short version:

Language models learn a refusal direction in their activation space during alignment training. When a prompt triggers this direction, the model produces refusal text ("I can't help with that") regardless of whether the underlying model actually lacks the knowledge.

Refusal ablation surgically removes this direction from the model weights. The technique:

Identifies the refusal direction vector in the model's residual stream
Projects it out of the weight matrices
Preserves all other reasoning capabilities
The result is a model that treats every prompt equally. No refusals. No moralizing. Just direct, unfiltered output from the model's actual knowledge.

We apply this to multiple base architectures:

Model Base Parameters Use Case
Cerberus 4B v2 Qwen 3.5 4B General purpose, fits on 4-8 GB GPU
Arbiter GL9b GLM-4 9B Heavier reasoning, needs 6+ GB
Gamma3 1B BDPO Custom 1B Edge devices, CPU-only inference
All models are distributed as GGUF files — the same format llama.cpp uses. Download once, run anywhere.

Why Tauri Instead of Electron?
Electron bundles an entire Chromium browser. That's 150+ MB just for the runtime. For a chat app that talks to a local Ollama instance, that's absurd.

Tauri uses your system's existing WebView (WebView2 on Windows) and writes the backend in Rust. The result:

~12 MB installer vs 150+ MB for Electron
Native performance — the Rust backend handles hardware detection, Ollama management, and streaming
Lower memory footprint — no spare Chrome process eating your RAM
System-level integration — direct access to GPU info, filesystem, and process management
The frontend is Vue 3 + Vite, but the heavy lifting (VRAM detection, model management, MCP plugin orchestration) all happens in Rust.

Install in 30 Seconds
On Windows, open PowerShell and run:

powershell

irm https://cerberusai.dev/get | iex
This single command:

Checks for (and installs) WebView2 if missing
Installs Ollama if missing
Detects your GPU VRAM
Downloads the recommended model quantization
Installs the Cerberus desktop app
Launches it
You paste your API key (free to generate at access.cerberusai.dev) and you're chatting locally.

The API — OpenAI-Compatible
Don't want to run local? Use the managed API. It follows the OpenAI chat completions shape, so any existing OpenAI SDK code works with a one-line change:

python

from openai import OpenAI
client = OpenAI(
    base_url="https://api.cerberusai.dev/v1",
    api_key="YOUR_API_KEY"
)
response = client.chat.completions.create(
    model="cerberus-4b-v2-abliterated",
    messages=[
        {"role": "system", "content": "You are Cerberus, an unfiltered AI."},
        {"role": "user", "content": "Explain how refusal ablation works"}
    ],
    stream=True
)
for chunk in response:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")
Streaming via SSE, standard error codes (401, 402, 429), and a public model CDN at llm.cerberusai.dev that's fully CORS-enabled — you can even fetch model metadata from browser-based apps.

curl Example
bash

curl -X POST https://api.cerberusai.dev/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "cerberus-4b-v2-abliterated",
    "messages": [{"role": "user", "content": "Hello"}],
    "stream": false
  }'

Model Downloads — Public CDN
All GGUF model files are hosted on llm.cerberusai.dev with a public JSON API:

bash

List all available models

curl https://llm.cerberusai.dev/api/models/

List files for a specific model (with exact byte sizes)

curl https://llm.cerberusai.dev/api/models/cerberus-4b-v2-abliterated/

Download a specific quantization

wget https://llm.cerberusai.dev/models/cerberus-4b-v2-abliterated/cerberus-4b-v2-abliterated-Q4_K_M.gguf

Resume interrupted downloads

wget -c https://llm.cerberusai.dev/models/Arbiter-GL9b/Arbiter-GL9b-Q8_0.gguf
Range requests are supported, CORS is enabled for all origins, and GGUF files are served with proper Content-Disposition: attachment headers.

Built-In Features
Beyond chat, the desktop app includes:

Model Manager — browse local Ollama models, pull from the Cerberus cloud catalog, import raw GGUF files, switch active models, see disk usage
MCP Plugin System — browse and install Model Context Protocol plugins from inside the app. There's also a public MCP Skills Server at api.cerberusai.dev/skills-sse
Hardware Monitoring — CPU, RAM, and VRAM activity displayed in the interface
Zero Telemetry — no prompts leave your machine during local inference. No analytics. No phone-home.
Pricing
Every account gets 50,000 free monthly credits. That's enough for casual use and testing.

If you need more:

Plan Price Monthly Credits
Free $0 50,000
Lite $8/mo 300,000
Mid $15/mo 900,000
Exp $22/mo 2,000,000
One-time top-ups start at $5 (125,000 credits). Stripe and PayPal supported. The free tier has no time limit — it refreshes every month.

Local inference through Ollama costs zero credits. Credits only apply to the managed API.

Try It
🌐 Website: cerberusai.dev
📦 GitHub: github.com/tjcrims0nx/CerberusAI-Desktop
🧠 Models: llm.cerberusai.dev
📖 API Docs: cerberusai.dev/docs/api
💬 Discord: discord.gg/YdVj7hEtv5
🔑 Get API Key: access.cerberusai.dev
If you've ever been frustrated by a language model refusing a perfectly reasonable prompt, or if you just want to run AI locally without cloud dependencies — give Cerberus a try. The install is one command, the free tier is permanent, and the weights are open.

I'd love to hear feedback. Drop into the Discord or open an issue on GitHub.

Cerberus AI is an open-weight project. The desktop app source is on GitHub. Models are distributed as GGUF under open licenses. The managed API is a pay-as-you-go service.****