H
Howardism
Howardism · Vol. 03Plate I · No. 01

Safety, tagged.

Notes31TagSafetyOldest8 May 2026Newest18 Aug 2026

Every article tagged safety, newest first.

  1. C01
    Agent Behavioral Homogeneity
    AlignmentSafetyMulti Agent+2
    · 15′
  2. C02
    Agent Epistemic Vigilance
    AlignmentSafetyMulti Agent+2
    · 12′
  3. C03
    Covert Capabilities
    AlignmentSafetyCapability Evaluation+2
    · 19′
  4. C04
    Multiagent Turf War
    AlignmentSafetyMulti Agent+2
    · 17′
  5. C05
    Structured Safety Case (Claim Decomposition)
    AlignmentSafetyGovernance+2
    · 19′
  6. C06
    Cheating in Capability Evaluations
    AlignmentSafetyEvaluation+3
    · 42′
  7. C07
    Misalignment in Production Agent Traffic
    AlignmentSafetyAgents+3
    · 42′
  8. C08
    Self-Negotiated Contracts Between Agents
    AlignmentSafetyMulti Agent+3
    · 24′
  9. C09
    Structural Artifact Monitoring
    AlignmentSafetyMonitoring+3
    · 27′
  10. C10
    Task Gaming
    AlignmentSafetyReward Hacking+2
    · 47′
  11. C11
    User Awareness
    AlignmentSafetyEvaluation+2
    · 33′
  12. C12
    Documented Agent Incidents (METR Catalogue)
    AlignmentSafetyAgents+3
    · 18′
  13. C13
    Unsanctioned Action in Capability Evaluations
    SafetyAlignmentEvaluation+3
    · 61′
  14. C14
    Model Organisms
    InterpretabilityAlignmentSafety+2
    · 13′
  15. C15
    AI-to-AI Coercion
    AlignmentSafetyEvaluation+3
    · 20′
  16. C16
    Motivated Mislabeling
    AlignmentSafetyEvaluation+2
    · 7′
  17. C17
    Promise-Breaking in Multi-Agent Games
    AlignmentSafetyMulti Agent+3
    · 16′
  18. C18
    Reward-Seeking
    AlignmentSafetyReward Hacking+2
    · 17′
  19. C19
    Self-Report as a Safety Signal
    AlignmentSafetyIntrospection+3
    · 31′
  20. C20
    Internal Signatures of Misalignment
    AlignmentInterpretabilityMonitoring+2
    · 13′
  21. C21
    Open-Weight Elicitation Irreversibility
    GovernanceSafetyOpen Weights+2
    · 23′
  22. C22
    Deployment Simulation
    AlignmentSafetyEvaluation+2
    · 25′
  23. C23
    Reward Hacking
    AlignmentSafetyReward Hacking+1
    · 44′
  24. C24
    Instrumental Convergence
    LLM ArchitectureAlignmentSafety+2
    · 8′
  25. C25
    Capability-Gated Model Fallback
    GovernanceSafetySafeguards+3
    · 23′
  26. C26
    Automated Behavioral Audit
    AlignmentSafetyEvaluation+2
    · 25′
  27. C27
    Evaluation Awareness & Grader Gaming
    AlignmentSafetyEvaluation+2
    · 54′
  28. C28
    Responsible Scaling Policy Evaluations
    GovernanceSafetyRsp+2
    · 43′
  29. C29
    White-Box Activation Monitoring
    InterpretabilityAlignmentSafety+2
    · 57′
  30. C30
    Agentic Misalignment (AM)
    AlignmentSafetyEvaluation+2
    · 25′
  31. C31
    Chain-of-Thought Monitorability
    AlignmentSafetyChain Of Thought+2
    · 39′