Search papers, labs, and topics across Lattice.
4
0
8
7
LLMs can sift through routine clinical notes to detect epilepsy with high accuracy, even boosting expert neurologists' diagnostic performance by over 10%.
Biomedical language models suffer severe catastrophic forgetting when sequentially updated, but parameter isolation offers the best retention per GPU-hour, revealing a crucial efficiency-stability tradeoff.
Clinicians using HeartAgent, a cardiology-specific agent system, improved diagnostic accuracy by 26.9% and explanatory quality by 22.7% compared to unaided experts.
Why waste tokens on reasoning when you don't need to? Selective Chain-of-Thought cuts LLM inference costs by up to 47% in medical QA, with minimal accuracy loss.
