3 papers
cs.CL2026
Consistency Training while Mitigating Obfuscation via Rate Matching
Sohaib Imran, Prakhar Gupta, Jannes Elstner +1
Large language models are often influenced by extraneous input features, such as cues revealing a user's preferred answer. Consistency training reduces this influence by training m…
cs.AI2026
Evaluating and Understanding Scheming Propensity in LLM Agents
Mia Hopman, Jannes Elstner, Maria Avramidou +2
As frontier language models are increasingly deployed as autonomous agents pursuing complex, long-term objectives, there is increased risk of scheming: agents covertly pursuing mis…
cs.LG2026
The Geometry of Refusal in Large Language Models: Concept Cones and Representational Independence
Tom Wollschläger, Jannes Elstner, Simon Geisler +3
The safety alignment of large language models (LLMs) can be circumvented through adversarially crafted inputs, yet the mechanisms by which these attacks bypass safety barriers rema…