Does Steering Break Your Model? A Multi-Dimensional Evaluation Suite for LLM Steering Methods
cs.CL
Submitted: 2026-10-06
Updated: 2026-10-06
Code: https://github.com/yht0511/steerscope
Terminology
Sources
- Towards Unifying Interpretability and Control: Evaluation via Intervention
- Conditional Optimal Bridge for Riemannian Activation Steering
- DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
- FaithSteer-BENCH: A Deployment-Aligned Stress-Testing Benchmark for Inference-Time Steering
- Gemma 2: Improving Open Language Models at a Practical Size
- ORBIT: Training-Free Multi-Attribute Behavioral Steering via Orthogonal Subspace Rotation
- The Llama 3 Herd of Models
- Contextual Linear Activation Steering of Language Models
- Global Evolutionary Steering: Refining Activation Steering Control via Cross-Layer Consistency
- Beyond Steering Vector: Flow-based Activation Steering for Inference-Time Intervention
- Manifold-Guided Attention Steering
- Steering Vector Fields for Context-Aware Inference-Time Control in Large Language Models
- Beyond Linear Activation Steering: Invertible Latent Transformations for Controlling LLM Behavior
- High-Dimensional Random Projection for Activation Steering in Language Models
- Curveball Steering: The Right Direction To Steer Isn't Always Linear
- UniSteer: Text-Guided Flow Matching in Activation Space for Versatile LLM Steering
- Mind the Performance Gap: Capability-Behavior Trade-offs in Feature Steering
- HyperSteer: Activation Steering at Scale with Hypernetworks
- Steering Language Models With Activation Engineering
- Instruction-Following Evaluation for Large Language Models
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering