Search papers, labs, and topics across Lattice.
The Hong Kong University of Science and Technology (Guangzhou
4
0
7
Static benchmarks mislead model performance assessments, as LiveHouse-TS reveals dramatic shifts in rankings when evaluated in real-time environments.
Exploiting client heterogeneity as complementary observations leads to a breakthrough in spatio-temporal forecasting, outperforming traditional methods.
Forget complex memory architectures: simple retrieval and generation, when carefully tuned for signal density, can outperform sophisticated methods in conversational agents.
Time series foundation models in federated learning can be significantly improved by representing data as discrete prototypical memories, rather than relying on continuous latent spaces.
