Search papers, labs, and topics across Lattice.
3
0
5
Across answer-focused, mixed-reasoning, and CoT-sensitive benchmarks, routed control improves robustness over fixed long decoding, pure short decoding, and single-rule interventions.
Current Personalized Large Language Models falter when user profiles and preferences diverge, revealing a critical gap in their reasoning capabilities.
LLMs exhibit alarming levels of stigma expression in group dialogues, with detection rates plummeting and unrealistic advice proliferating under social pressure.
