Skip to content

Scan record

Everything the site collected, newest first by collection time (Beijing time), back seven days.

Nov 1740 loaded

  1. Between Rigor and Reality: How AI Safety Benchmark Developers Understand Benchmark Use and Maintenance

    PaperarXiv: Alignment, Deception & OversightThe item you came from

  2. Manipulating Bayesian inference and decision-making through data poisoning

    PaperarXiv: 其他 AI 安全论文

  3. Robustness as a Transferable Skill: Cross-Task Adversarial Defense for Vision-Language-Action Policies

    PaperarXiv: 其他 AI 安全论文

  4. The Paradox of Provenance and Privacy: Can Private Cloud Compute (PCC) Prove That Apple Reference Images Are Not Used for Learning?

    PaperarXiv: 其他 AI 安全论文

  5. When tool names are not identity: measuring provider misbinding in MCP-based AI agents

    PaperarXiv: 其他 AI 安全论文

  6. Optimistic verifiers get gamed: reward hacking under exact process oracles

    PaperarXiv: 其他 AI 安全论文

  7. Similar in Safety, Silent in Propagation: Limitations of Text-Similarity Guardrails for AI Agent Security

    PaperarXiv: 其他 AI 安全论文

  8. Evaluating Authorization and Trust Boundaries in Agent Orchestration Frameworks

    PaperarXiv: 其他 AI 安全论文

  9. Jev-as-a-Reward: Validating a Probabilistic Safety Judge for Reinforcement Learning

    PaperarXiv: 其他 AI 安全论文

  10. Experience Rot: Experienced agents fail where novices succeed

    PaperarXiv: 其他 AI 安全论文

  11. Which AI Use Cases Are Worth Piloting? A Risk and Value Framework for Funding, Scoping and Scaling AI

    PaperarXiv: 其他 AI 安全论文

  12. Tight PAC-Bayes Generalization Guarantees for LLM Judges

    PaperarXiv: 其他 AI 安全论文

  13. STA: A Split Temporal Adversarial Method for Evaluating the Robustness of Multimodal Reinforcement Learning Driving Policies

    PaperarXiv: 其他 AI 安全论文

  14. LEAF-OPD: Learning to Unlearn from Privileged Context via On-Policy Distillation

    PaperarXiv: 其他 AI 安全论文

  15. Who Bears the Excess Authority? Attributing AI Agent Behavior to the Delegating Party

    PaperarXiv: 其他 AI 安全论文

  16. Safety-Preserving Fine-Tuning for Medical LLMs via Bi-Level Data Selection

    PaperarXiv: 其他 AI 安全论文

  17. Counterfactual feature steering for identity unlearning with adjacency preservation

    PaperarXiv: 其他 AI 安全论文

  18. Example-based spatial guidance: training-free concept erasure in diffusion models

    PaperarXiv: 其他 AI 安全论文

  19. GPromptShield: Enhancing the Resilience of Graph Prompt Tuning under Adversarial Attacks

    PaperarXiv: 其他 AI 安全论文

  20. SmoothShift: Disrupting Visual Adversarial Attacks against LVLMs at Test Time

    PaperarXiv: 其他 AI 安全论文

  21. Model Immunity Is More Than Condition Number: The Importance of Plateau Regions

    PaperarXiv: 其他 AI 安全论文

  22. Designing Deceptive Model Organisms That Successfully Evade Auditing

    PaperarXiv: 其他 AI 安全论文

  23. Interpretable Federated Learning via Global-Local Attribution Alignment

    PaperarXiv: 其他 AI 安全论文

  24. Abstention directions in LLMs: mechanistic interpretability of abstention in query ambiguity, knowledge boundaries, and safety alignment

    PaperarXiv: 其他 AI 安全论文

  25. Refusal Register Can Be Causally Manipulated: Decomposing the Refusal Direction of Gemma-2-9B-IT into Category-Aligned Components

    PaperarXiv: 其他 AI 安全论文

  26. Where the Evidence Hides: Measuring Privacy and Diagnosability of Exported Agent Trajectories

    PaperarXiv: 其他 AI 安全论文

  27. Where the saved time goes: limitations of hiding cross-tenant cache hits in LLM serving

    PaperarXiv: 其他 AI 安全论文

  28. Protecting safety in reasoning training with low-rank Fisher regularization

    PaperarXiv: 其他 AI 安全论文

  29. Who Can You Hide Behind? Author-Conditioned Style Transfer and Evasion of a Commercial AI-Text Detector

    PaperarXiv: 其他 AI 安全论文

  30. Risk-Certified Prompt Optimization for LLM-as-a-Judge Guardrails

    PaperarXiv: 其他 AI 安全论文

  31. HEAMS: Holistic Evaluation of Agent Memory Systems

    PaperarXiv: 其他 AI 安全论文

  32. Can AI judge you? An empirical audit of the deployment readiness of LLM-assisted hiring systems

    PaperarXiv: 其他 AI 安全论文

  33. "Don't leave the task unfinished": persistence-driven violations under silent supervision

    PaperarXiv: 其他 AI 安全论文

  34. HalluProbe: using VLM hallucinations for backdoor detection

    PaperarXiv: 其他 AI 安全论文

  35. Where in multimodal LM-derived models can Rank-1 DiM refusal be intervened on?

    PaperarXiv: 其他 AI 安全论文

  36. Fine-Grained RLAIF for Visual Safety: Learning to Magnify Hazards for Multimodal Safety Alignment

    PaperarXiv: 其他 AI 安全论文

  37. When backdoors can be switched off: suppression-aware backdoor attacks on vision-language-action models

    PaperarXiv: 其他 AI 安全论文

  38. Building enterprise AI infrastructure: the evolution of the AI governance committee and the chief ethics officer role

    PaperarXiv: 其他 AI 安全论文

  39. BadGCL: Persistent Backdoor Attacks Against Graph Continual Learning Systems

    PaperarXiv: 其他 AI 安全论文

  40. Protecting the system while leaving users stranded: the limits of multi-agent authorization

    PaperarXiv: 其他 AI 安全论文