System Prompts Affect Response Bias More Than Discriminability in LLM Safety
cs.LG, cs.AI, cs.CL, cs.SE
Submitted: 2026-11-08
Updated: 2026-11-08
Terminology
Sources
- Refusal in Language Models Is Mediated by a Single Direction
- LLMs as Signal Detectors: Sensitivity, Bias, and the Temperature-Criterion Analogy
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- The Llama 3 Herd of Models
- Gemma 4 Technical Report
- Do Language Models Know When They'll Refuse? Probing Introspective Awareness of Safety Boundaries
- In-Context Learning Creates Task Vectors
- A Closer Look at System Prompt Robustness
- Qwen2.5 Technical Report
- Function Vectors in Large Language Models
- The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions
- An Explanation of In-context Learning as Implicit Bayesian Inference
- Qwen2 Technical Report
- IHEval: Evaluating Language Models on Following the Instruction Hierarchy
- Universal and Transferable Adversarial Attacks on Aligned Language Models
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