3 papers
cs.LG2026
Genesis Steering: Better Activation Steering Through Training Data
Tora Bodin, Eric Aubinais, Ilaria Torre +1
Activation steering controls large language models at inference time, but its effectiveness depends on the choice of reference data. We introduce genesis steering, which derives a…
cs.AI2026
A Cross-Architecture Audit of Direction-Based Inference-Time Defences in Vision-Language Models
Xiangyu Yin, Tora Bodin, Rohan Menon +1
Inference time defences against vision language model jailbreaks often subtract a calibrated direction from the residual stream at a chosen decoder layer. We compare five defence c…
cs.RO2025
An integrated process for design and control of lunar robotics using AI and simulation
Daniel Lindmark, Jonas Andersson, Kenneth Bodin +4
We envision an integrated process for developing lunar construction equipment, where physical design and control are explored in parallel. In this paper, we describe a technical fr…