Search papers, labs, and topics across Lattice.
1
0
3
1
Fine-tuning LLMs on social media data can improve the superficial style of generated comments but simultaneously degrade their semantic accuracy, revealing a critical disconnect between form and content.
