Search papers, labs, and topics across Lattice.
Affiliation:
3
0
6
Stacked, a novel supervised method, significantly enhances hallucination detection in LLMs, outperforming alternatives in nearly half of the tested scenarios.
LLMs' apparent moral neutrality is a mirage: even when they claim impartiality, subtle contextual cues can dramatically sway their decisions in moral dilemmas, sometimes even in the opposite direction of the cue.
