Abliteration
Abliteration is the practice of removing refusal behavior from an open-weight large language model by editing its weights so that the model can no longer write a "refusal direction" into its residual stream.
Open Source AI
Explore Open Source AI through related topics and the articles other pages reference most.
Explore articles
Browse subtopics: AI Safety
Articles that also belong to these categories. Counts cover all of Open Source AI.
All articles (473)Large Language Models (173)AI Models (123)Developer Tools (117)Chinese AI (87)AI Agents (60)AI Companies (47)AI Infrastructure (43)Mixture of Experts (42)Robotics (31)Multimodal AI (29)NVIDIA (29)AI Code Generation (25)Machine Learning (25)Natural Language Processing (23)Generative AI (18)AI Tools & Products (17)Meta AI (17)Small Language Models (17)AI Inference (15)Data & Datasets (15)Diffusion Models (14)Image Generation (13)Information Retrieval (13)Video Generation (13)Computer Vision (12)Google (11)Reasoning Models (11)AI Research (10)Embodied AI (10)Research Organizations (10)AI Hardware (9)AI Safety (9)Artificial Intelligence (9)Microsoft (9)Speech & Audio AI (9)Google DeepMind (8)MLOps (7)Reinforcement Learning (7)Training & Optimization (7)AI for Science (6)OpenAI (6)Robot Hardware (6)Deep Learning (5)Drug Discovery (5)Voice AI (5)AI Benchmarks (4)Enterprise AI (4)Healthcare AI (4)Model Evaluation (4)Algorithms (3)Conversational AI (3)Humanoid Robots (3)Interpretability (3)People (3)AI Ethics (2)AI Policy & Regulation (2)Agentic Commerce (2)Anthropic (2)Mathematics (2)Model Architecture (2)Neural Networks (2)Physical AI (2)Programming Languages (2)Software Development (2)Transformer Models (2)AI Events (1)AI Incidents & Controversies (1)AI Policy (1)Data Science (1)Music & Audio Generation (1)Quantum Computing (1)Robotics Companies (1)World Models (1)
Showing 1-9 of 9 articles
Heretic (software)
Heretic is an open-source command-line program that removes refusal behavior from transformer-based language models without retraining them.
NVIDIA Open Agent Safety Platform
The NVIDIA Open Agent Safety Platform is an open software platform and reference system design from NVIDIA for containing and monitoring autonomous AI agents.
NVIDIA OpenShell
NVIDIA OpenShell is an open source runtime that executes autonomous AI agents inside policy-governed sandboxes, published by NVIDIA under the Apache License 2.0.
Open Secure AI Alliance
The Open Secure AI Alliance is an industry coalition announced on July 27, 2026 to develop and share open technologies, techniques, and tools for securing software and AI agents.
OpenClaw
OpenClaw is a free, open-source autonomous AI agent that runs on a user's own machine or server and executes real-world tasks through everyday messaging apps such as WhatsApp, Telegram, Discord, Slack, and…
Petri (auditing tool)
Petri is an open-source tool, released by Anthropic, that automates parts of AI alignment auditing by using one AI model to probe another and a third to score what happens.
Purple Llama
Purple Llama is an umbrella project from Meta that bundles open, permissively licensed trust and safety tools and evaluations for generative AI.
ShieldGemma
ShieldGemma is a family of open safety-classifier models from Google, built on the Gemma family of lightweight open models.