AI RESEARCH

Language-Switching Triggers Take a Latent Detour Through Language Models

arXiv CS.CL

ArXi:2605.18646v1 Announce Type: new Backdoor attacks on language models pose a growing security concern, yet the internal mechanisms by which a trigger sequence hijacks model computations remain poorly understood. We identify a circuit underlying a language-switching backdoor in an 8B-parameter autoregressive language model, where a three-word Latin trigger (nine tokens) redirects English output to French.