WhitzardNUWA Lab ↗
NUWANUWA INDEX

Search NUWA Index

Research · Benchmarks · Findings · Perspectives

中文
← Library

Institution

MATS

Roles and affiliations follow the cited source. A source check is not an independent replication.

More filters →

MATS ↗

Multi-Agent AI Control: Distributed Attacks Hamper Per-Instance Monitors

QuestionCan distributed coordination evade monitors that inspect one agent’s commit at a time?

Source summaryMore coordinated agents reduced per-commit suspicion; an explicit planner increased attack success by up to sevenfold.

NUWA research note · independent review pending · Contains: question / method / findings / interpretation / limits

Ctrl-Z: Controlling AI Agents via Resampling

QuestionCan a multi-step agent be kept useful while stopping a deliberately subversive policy under a limited audit budget?

Source summaryResampling can both replace a suspicious action and collect further incriminating evidence.

NUWA research note · independent review pending · Contains: question / method / findings / interpretation / limits / connections