Author

Jérémie Decouchant

h-index 4 View on Semantic Scholar

We have 2 of 31 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Preprint Sep 2026

Absorbed in Inertia: Activation Analysis for Computer-Use Agents

This work proposes R$^3$ (Reset, Reroute, Restore), which temporarily resets the agent's context trajectory to escape the absorbing region and then restores the historical context to effectively complete the task, suggesting that changing the context can interrupt recurrence more effectively than directly steering the...

Giulio Segalini, Wen-Soi Zhi, Jérémie Decouchant et al. · 0 citations

Preprint Aug 2026

Diffusion LLMs as Targets and Adversaries: Mechanistic Safety Exploits

SN-Guided Diffusion is introduced, a fully offline black-box jailbreak framework that steers the diffusion process away from safety-triggering regions using a weighted safety neuron loss, which achieves near-perfect prompt separability.

Elena Dumitrescu, Gert Lek, L. Chen et al. · 0 citations