22 papers
No Safe Dose: How Training Data Drives Unsafe Image Generation
Felix Friedrich, Lukas Helff, Niharika Hegde +2
Text-to-image models trained on large-scale data often inevitably ingest unsafe content. While some people observe input-output amplifications, it remains unclear whether and how t…
Multilingual Steering by Design: Multilingual Sparse Autoencoders and Principled Layer Selection
Yusser Al Ghussin, Daniil Gurgurov, Tanja Baeumel +3
Sparse autoencoders (SAEs) enable feature-level mechanistic interpretability and activation steering in large language models (LLMs), but SAE-based language control remains unrelia…
ActivationReasoning: Logical Reasoning in Latent Activation Spaces
Lukas Helff, Ruben Härle, Wolfgang Stammer +6
Large language models (LLMs) excel at generating fluent text, but their internal reasoning remains opaque and difficult to control. Sparse autoencoders (SAEs) make hidden activatio…
SLR: Automated Synthesis for Scalable Logical Reasoning
Lukas Helff, Ahmad Omar, Felix Friedrich +7
We introduce SLR, an end-to-end framework for systematic evaluation and training of Large Language Models (LLMs) via Scalable Logical Reasoning. Given a user's task specification,…
SocialGrid: A Benchmark for Planning and Social Reasoning in Embodied Multi-Agent Systems
Hikaru Shindo, Hanzhao Lin, Lukas Helff +2
As Large Language Models (LLMs) transition from text processors to autonomous agents, evaluating their social reasoning in embodied multi-agent settings becomes critical. We introd…
LLMs Gaming Verifiers: RLVR can Lead to Reward Hacking
Lukas Helff, Quentin Delfosse, David Steinmann +6
As reinforcement Learning with Verifiable Rewards (RLVR) has become the dominant paradigm for scaling reasoning capabilities in LLMs, a new failure mode emerges: LLMs gaming verifi…