Stress-Testing Sandbagging Detection Under Adaptive Attack
cs.CR, cs.LG
Submitted: 2026-11-11
Updated: 2026-11-11
Code: https://github.com/divyashie/ai-sandbagging
Terminology
Sources
- The Internal State of an LLM Knows When It's Lying
- Linear probes rely on textual evidence: Results from leakage mitigation studies in language models
- Training Verifiers to Solve Math Word Problems
- The Llama 3 Herd of Models
- RL-Obfuscation: Can Language Models Learn to Evade Latent-Space Monitors?
- LoRA: Low-Rank Adaptation of Large Language Models
- Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
- Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense
- Neural Chameleons: Language Models Can Learn to Hide Their Thoughts from Unseen Activation Monitors
- Qwen2.5 Technical Report
- Can AI-Generated Text be Reliably Detected?
- Auditing Games for Sandbagging
Related papers
- From Strategy Search to Policy Learning: MCTS-Guided Adversarial Co-Evolution for LLM Safety Alignment
- Privacy Leakage from Recurrent State in LLM Inference
- Incrementally Verifiable Oracle Aided Computation
- Privilege and confidentiality in generative AI workflows
- BadMoE: Backdooring Mixture-of-Experts LLMs via Routing-Aware Triggers and Task-Underutilized Experts
- CoVault: Secure, Scalable Analytics of Personal Data