Language-Switching Triggers Take a Latent Detour Through Language Models
Francis Kulumba Wissam Antoun Théo Lasnier, Benoît Sagot Djamé Seddah · 2026
Read Language-Switching Triggers Take a Latent Detour Through Language Models by Francis Kulumba Wissam Antoun Théo Lasnier, Benoît Sagot Djamé Seddah (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-18646.txt.
Description
Francis Kulumba and four co-authors identify, in a 2026 paper, the circuit behind a language-switching backdoor in an 8B-parameter autoregressive language model. A three-word Latin trigger redirects English output to French. Early attention heads compose trigger tokens at the last position. The signal then travels mid-layers in a subspace orthogonal to the natural language-identity direction. The final-layer MLP converts it into French logits. Corrupting the single-position bottleneck at any layer kills the trigger but also harms capabilities.
Catalog record
| CREATOR | Francis Kulumba Wissam Antoun Théo Lasnier, Benoît Sagot Djamé Seddah |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47475 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.18646