Search papers, labs, and topics across Lattice.
4
0
5
2
Answer format can dramatically alter gender bias measurements in LLMs, sometimes reversing bias rankings entirely.
Stance shifts in LLMs can be systematically induced by linguistic constructions, revealing critical vulnerabilities in their decision-making processes.
Transformer language models stumble on complex syntactic structures, failing to mimic human-like error patterns in agreement attraction, suggesting current architectures lack crucial aspects of human morphosyntactic processing.
LLMs are surprisingly good at making sense of nonsense, generating plausible contexts even for sentences designed to be semantically deviant.
