4 papers
Understanding Confabulation and Rethinking Reconstruction in Activation Explanations
Gert Lek, Zixuan Xia, Pin-Yu Chen +1
Natural Language Autoencoders (NLAs) produce unsupervised text explanations of a model's activations: a verbalizer describes an activation and a reconstructor learns to recover it…
Safety Reconstructed: Generative Modeling via Masked Diffusion Builds Strong Safety Guardrails
Gert Lek, Abele Malan, Chaoyi Zhu +3
Guard models are the last line of defense between a language model and a harmful output, yet their training objective is surprisingly narrow. Existing guards learn to predict a sin…
Diffusion LLMs as Targets and Adversaries: Mechanistic Safety Exploits
Elena Dumitrescu, Gert Lek, Lydia Y. Chen +1
Diffusion Large Language Models (DLLMs) replace autoregressive next-token prediction with iterative parallel denoising, yet their internal safety mechanisms remain poorly understoo…
Detecting Diffusion-Generated Time Series Under Generator Shift
Zhi Wen Soi, Aditya Shankar, Gert Lek +4
The boundary between real and diffusion-generated time series is becoming increasingly difficult to draw, yet detection in this domain remains underexplored, especially when the ge…