Search papers, labs, and topics across Lattice.
2
3
6
13
LLM safety can be significantly bolstered against harmful fine-tuning attacks by strategically projecting models back into safe parameter space using a relevance- and diversity-aware curated dataset.
Stop feeding VLMs the same graph representation for every question – DynamicGTR dynamically picks the best one, boosting accuracy and brevity in graph Q&A without retraining.
