Search papers, labs, and topics across Lattice.
3
0
6
2
Predicting task-relevant latent futures, not raw pixels, unlocks efficient and effective long-horizon VLA control, outperforming reactive policies by over 11 percentage points.
A single verification prompt is insufficient; learning a reliability score from multiple, complementary perspectives unlocks robust self-verification for VLMs.
Extend your LLM's context window by 16x without retraining from scratch: just "self-inject" a compressed representation from a smaller model into a larger one.
