World models introduce a stealthy poisoning vector into robot learning pipelines where malicious prompts or dynamics in teleoperated data activate only during synthetic trajectory generation, enabling backdoors in downstream policies.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
MAVIC corrects Bellman backups at instruction boundaries by adjusting the incoming objective and restoring continuation value, enabling consistent estimation under stochastic instruction switching in cooperative MARL.
citing papers explorer
-
Targeting World Models to Compromise Robot Learning Pipelines
World models introduce a stealthy poisoning vector into robot learning pipelines where malicious prompts or dynamics in teleoperated data activate only during synthetic trajectory generation, enabling backdoors in downstream policies.
-
Robust Instruction Compliance in Cooperative Multi-Agent Reinforcement Learning
MAVIC corrects Bellman backups at instruction boundaries by adjusting the incoming objective and restoring continuation value, enabling consistent estimation under stochastic instruction switching in cooperative MARL.