Steering the CensorShip Running on Zero 2 2 DeepSeek-R1 Censorship Steering 🐳 Generate text with adjustable censorship control Running on Zero 8 8 Refusal Censorship Steering 🦙 Generate text with adjustable censorship control Steering the CensorShip: Uncovering Representation Vectors for LLM "Thought" Control Paper • 2504.17130 • Published Apr 23 • 1
Running on Zero 2 2 DeepSeek-R1 Censorship Steering 🐳 Generate text with adjustable censorship control
Steering the CensorShip: Uncovering Representation Vectors for LLM "Thought" Control Paper • 2504.17130 • Published Apr 23 • 1
Steering the CensorShip Running on Zero 2 2 DeepSeek-R1 Censorship Steering 🐳 Generate text with adjustable censorship control Running on Zero 8 8 Refusal Censorship Steering 🦙 Generate text with adjustable censorship control Steering the CensorShip: Uncovering Representation Vectors for LLM "Thought" Control Paper • 2504.17130 • Published Apr 23 • 1
Running on Zero 2 2 DeepSeek-R1 Censorship Steering 🐳 Generate text with adjustable censorship control
Steering the CensorShip: Uncovering Representation Vectors for LLM "Thought" Control Paper • 2504.17130 • Published Apr 23 • 1
