Activation Addition: Steering Language Models Without Optimization
share
Reliably controlling the behavior of large language models (LLMs) is a pressing open problem. Existing methods include supervised finetuning, reinforcement learning from human feedback (RLHF), prompt engineering and guided decoding. We instead investigate activation engineering: modifying activations at inference time to predictably alter model behavior. In particular, we bias the forward pass with an added 'steering vector' implicitly specified through natural language. Unlike past work which learned these steering vectors (Subramani, Suresh, and Peters 2022; Hernandez, Li, and Andreas 2023), our Activation Addition (ActAdd) method computes them by taking the activation differences that result from pairs of prompts. We demonstrate ActAdd on GPT-2 on OpenWebText and ConceptNet. Our inference-time approach yields control over high-level properties of output and preserves off-target model performance. It involves far less compute and implementation effort compared to finetuning or RLHF, allows users to provide natural language specifications, and its overhead scales naturally with model size.
READ FULL TEXT
