Karolina Stanczak

Activations, not just descriptions, unlock the prediction of how LLM alignment generalizes to unseen values, revealing a structured value space.

Early layers of language models capture human-like processing signatures in reading, rivaling traditional measures like surprisal in predicting initial eye movements.

SLMs are shockingly vulnerable: combining adversarial audio and text unlocks 1.5x to 10x higher jailbreak rates than attacking either modality alone.