Search papers, labs, and topics across Lattice.
3
0
5
0
SIMAX reveals that AI-generated clinical dialogues can achieve high realism and fidelity, paving the way for scalable communication coding solutions in healthcare.
LLMs can denoise sequential recommendations by disagreeing with the recommendation model itself, leading to more robust performance against noisy user data.
By disentangling shared account behavior in the frequency domain, DisenReason dynamically infers latent users, boosting recommendation accuracy by up to 12.56% compared to methods assuming a fixed number of users.
