Reward Stealing Attack on Large Language Models
cs.CL
Submitted: 2026-10-05
Updated: 2026-10-05
Code: https://github.com/GarminQ/ReSA
Sources
- The Ethics of Interaction: Mitigating Security Threats in LLMs
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- Universal and Transferable Adversarial Attacks on Aligned Language Models
- Maximum Entropy Deep Inverse Reinforcement Learning
- Aligning Language Models with Preferences through f-divergence Minimization
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Shadow Alignment: The Ease of Subverting Safely-Aligned Language Models
- The Llama 3 Herd of Models
- Gemma: Open Models Based on Gemini Research and Technology
- Qwen2 Technical Report
- Baseline Defenses for Adversarial Attacks Against Aligned Language Models
- Safe RLHF: Safe Reinforcement Learning from Human Feedback
- Tulu 3: Pushing Frontiers in Open Language Model Post-Training
- FlipAttack: Jailbreak LLMs via Flipping
- Confidence Elicitation: A New Attack Vector for Large Language Models
- HarmRLVR: Weaponizing Verifiable Rewards for Harmful LLM Alignment
- ARGS: Alignment as Reward-Guided Search
- Cascade Reward Sampling for Efficient Decoding-Time Alignment
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