Search papers, labs, and topics across Lattice.
2
0
5
Hierarchical JEPAs unlock efficient long-horizon visual planning by learning distinct latent spaces for different temporal scales, dramatically boosting success rates and reducing compute.
End-to-end learned subspace projections shatter prior compression limits, achieving 70% weight reduction in Llama-2-7B with near-dense performance and drastically shrinking KV cache memory.
