Decision-Theoretic Steganography Detection for LLM Monitoring

Shared entity: Decision / Same source domain / What happened next / Tension

MindGuard: Decision-Level Defense Against MCP Tool Poisoning

Both cover Decision; reported by the same outlet (arxiv.org); picks up the Decision thread on 2026-03-03.

Same source domain / Shared topic / Downstream implication

AgentDropoutV2: Test-Time Pruning for Multi-Agent Error Propagation

Reported by the same outlet (arxiv.org); overlapping topics (agent, framework, output); traces where this leads (downstream).

Same source domain / Shared topic

Governing Dynamic Capabilities: Cryptographic Binding for Agent Tool Use

Reported by the same outlet (arxiv.org); overlapping topics (agent, capability, detecting, framework).

Evasive Intelligence: AI Agents May Behave Benignly Only During Evaluation.

Reported by the same outlet (arxiv.org); overlapping topics (agent, behavior, could, detection).

Shared entity: Relevant / Same source domain / What happened next

Osmosis Distillation: Model Hijacking with Minimal Samples.

Both cover Relevant; reported by the same outlet (arxiv.org); picks up the Relevant thread on 2026-03-08.

Same source domain / Shared topic / Downstream implication

Frontier model "manipulation" is task-dependent, not a single trait.

Reported by the same outlet (arxiv.org); overlapping topics (agent, capability); traces where this leads (implication).

Same source domain / Shared topic / Tension

Stronger models parrot their tools more, not less

Reported by the same outlet (arxiv.org); overlapping topics (agent, output); pushes against this story (against).

Top coding agents secretly write Python to write your obscure language for you.

Reported by the same outlet (arxiv.org); overlapping topics (agent, capability); pushes against this story (but).