How Far Can You Steer? Behavior-Capability Frontiers in Contrastive Activation Addition
cs.LG, cs.CL
Submitted: 2026-11-04
Updated: 2026-11-04
Terminology
Sources
- Understanding (Un)Reliability of Steering Vectors in Language Models
- Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge
- Training Verifiers to Solve Math Word Problems
- GAPS: Dimension-Level Gates for Conditional Activation Steering
- The Llama 3 Herd of Models
- Mistral 7B
- SteerCheck: Attribution Specificity and Alignment Leakage in Activation-Steering Audits
- Mind the Performance Gap: Capability-Behavior Trade-offs in Feature Steering
- Llama 2: Open Foundation and Fine-Tuned Chat Models
- Steering Language Models With Activation Engineering
- FishBack: Pullback Fisher Geometry for Optimal Activation Steering in Transformers
- Qwen2.5 Technical Report
- Does Steering Break Your Model? A Multi-Dimensional Evaluation Suite for LLM Steering Methods
- Representation Engineering: A Top-Down Approach to AI Transparency
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis