Search papers, labs, and topics across Lattice.
4
0
7
Forget small-scale SWE datasets: OpenSWE delivers 45K executable environments for training software engineering agents, unlocking SOTA results and surprising out-of-domain gains.
LLMs can spot fake words in speech by recognizing common editing patterns, but this reliance on learned biases hinders generalization to new manipulation techniques.
Achieve zero-shot voice conversion competitive with methods requiring more data or training, using a simple, invertible linear method to disentangle speech content from speaker timbre.
Surgeons can now visualize pre-planned trocar placements directly on a patient's body using only an AR headset, eliminating the need for external sensors or markers.
