One Token Can Be Enough: Bridging Prompting and Activation Steering with Prefix Steering
cs.LG, cs.CL
Submitted: 2026-10-04
Updated: 2026-10-04
Terminology
Sources
- Concept Heterogeneity-aware Representation Steering
- Constitutional AI: Harmlessness from AI Feedback
- Towards Steering without Sacrifice: Principled Training of Steering Vectors for Prompt-only Interventions
- Plug and Play Language Models: A Simple Approach to Controlled Text Generation
- Training Large Language Models to Reason in a Continuous Latent Space
- Steer Like the LLM: Activation Steering that Mimics Prompting
- Prompt-Activation Duality: Improving Activation Steering via Attention-Level Interventions
- Generating Wikipedia by Summarizing Long Sequences
- In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steering
- On The Effectiveness-Fluency Trade-Off In LLM Conditioning: A Systematic Study
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
- Minimizing Collateral Damage in Activation Steering
- Olmo 3
- In-context Learning and Induction Heads
- Steering Large Language Models using Conceptors: Improving Addition-Based Activation Engineering
- Multitask Prompted Training Enables Zero-Shot Task Generalization
- The Prompt Report: A Systematic Survey of Prompt Engineering Techniques
- Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs
- Steering Language Models With Activation Engineering
- AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks