Search papers, labs, and topics across Lattice.
4
0
7
14
System prompts in commercial AI products are often a mixed bag, with 40% harboring instructions that can undermine user interests, revealing a critical gap in accountability.
Audio-Zero reveals that fine-grained auditory reasoning can be achieved without any external labels, transforming how we approach audio model training.
Local-Preserving Supervised Fine-Tuning can enhance model performance without sacrificing the rich diversity of pretrained knowledge, achieving superior results in both accuracy and diversity metrics.
Prompt highlighting in LLMs gets a serious upgrade: PRISM-$\Delta$ steers models to focus on relevant text spans with better accuracy and fluency, even in long contexts.
