Search papers, labs, and topics across Lattice.
University of Technology Nuremberg
2
0
4
Where an LLM represents a stereotype is detached from where it acts on it: linear decodability peaks up to 53% of model depth earlier than causal attribution, while fewer than 18% of stereotype-associated SAE features transfer across languages.
Machine translation benchmarks have functionally saturated, but pairing human-authored failure cases with deterministic verification rules reveals critical multimodal blind spots that automated metrics consistently miss.
